TWI702805B - System and method for guiding a machine capable of autonomous movement - Google Patents

Abstract

A system for guiding a machine that can move autonomously, including: a machine that can move autonomously, on which a rolling shutter camera is installed; and an optical communication device, which includes a light source configured to work in at least two types Mode, the at least two modes include a first mode and a second mode, and wherein, in the first mode, the property of the light emitted by the light source is controlled at the first frequency by a light source control signal having a first frequency Continuously changing to present stripes on the image of the light source obtained when the light source is photographed by the rolling shutter camera. In the second mode, the light source is exposed to the light source by the rolling shutter camera. The image of the light source obtained during shooting does not exhibit stripes or exhibits stripes different from the stripes in the first mode.

Description

System and method for guiding a machine capable of autonomous movement

The present invention relates to the guidance of a machine that can move autonomously, and more specifically to a system and method for guiding a machine that can move autonomously through an optical communication device.

For the guidance of autonomously moving machines (for example, drones), currently, GPS, IMU and other technologies are usually implemented, but these technologies have limited positioning accuracy. For example, GPS usually has an error of several meters or even tens of meters, and its signal propagation will be affected by the environment in which it is used, often resulting in delay errors. Therefore, the current navigation technology can usually only guide the drone to the vicinity of the target location (for example, within a radius of tens of meters near the target location), but it is difficult to guide the drone to a very precise target location.

In recent years, many manufacturers have considered using drones for cargo distribution. For example, an Amazon patent US9536216B1 introduces a UAV cargo delivery system, which is based on GPS and altimeter to navigate the UAV, and can perform remote manual assisted navigation through the UAV camera. However, the above-mentioned system cannot achieve precise navigation of the drone. In another scheme disclosed by Amazon, the drone is first guided to the destination near the destination through GPS, and then the drone will look for a unique "mark" in its field of view, which is placed by the customer at a good landing site The electronic identification welcome mat with a predetermined pattern. If the "mark" is found, the drone will use visual guidance to fly to the mark and drop the package. However, the above method requires the buyer to have a In the courtyard, unique "marks" are placed in the courtyard. Moreover, since the "mark" itself cannot be used to distinguish between different buyers, if there are multiple "marks" placed by multiple buyers near the destination, the drone cannot determine which "mark" to place the package on. "Place. Therefore, the above scheme is not applicable to people living in urban apartments.

Traditional two-dimensional codes can be used to identify different users, but the recognition distance of two-dimensional codes is very limited. For example, for a two-dimensional code, when a camera is used to scan it, the camera must usually be placed within a relatively close distance, which is usually only about 15 times the width of the two-dimensional code. For example, for a two-dimensional code with a width of 20 cm, a drone equipped with a camera needs to travel about 3 meters from the two-dimensional code to be able to recognize the two-dimensional code. Therefore, for long-distance recognition, the two-dimensional code cannot be realized, or a very large two-dimensional code must be customized, but this will increase the cost, and in many cases it is impossible due to various other restrictions (such as space limitations) Achieved. Moreover, when recognizing a two-dimensional code, the camera needs to be roughly photographing the two-dimensional code. If the deviation angle is too large, the recognition will not be possible.

CMOS imaging devices are currently widely used imaging devices. As shown in FIG. 1, they include an array of image sensitive units (also called image sensors) and some other elements. The image sensor array may be a photodiode array, and each image sensor corresponds to a pixel. Each column of image sensors corresponds to a column amplifier, and the output signal of the column amplifier is then sent to the A/D converter (ADC) for analog-to-digital conversion, and then output through the interface circuit. For any image sensor in the image sensor array, it is now cleared at the beginning of the exposure, and then after the exposure time has passed, the signal value is read out. CMOS imaging devices usually use rolling shutter imaging. In CMOS imaging devices, the data readout is serial, so clearing/exposure/readout can only be carried out in a row-by-row sequence similar to a pipeline, and processed in all rows of the image sensor array Synthesize it into one frame after completion image. Therefore, the entire CMOS image sensor array is actually exposed line by line (in some cases, the CMOS image sensor array may also be exposed to multiple lines at a time), which results in the inter-line exposure. There is a small time delay. Due to this small time delay, when the light source flickers at a certain frequency, some undesirable stripes will appear on the image captured by the CMOS imaging device, which affects the shooting effect.

People have found that in theory, the stripes on the image taken by the CMOS imaging device can be used to transmit information (similar to a barcode), and try to transmit as much information as possible through the stripes, but this usually requires the CMOS imaging device and the light source Try to be as close as possible, and preferably always at a roughly fixed distance, and also requires fine time synchronization, accurate identification of the boundaries of each stripe, accurate detection of the width of each stripe, etc. Therefore, its stability in practice And reliability is not satisfactory, nor has it been widely used. Moreover, this method is obviously not suitable for long-distance recognition by a traveling drone.

One aspect of the present invention relates to a system for guiding a machine that can move autonomously, including: a machine that can move autonomously on which a rolling shutter camera is mounted; and an optical communication device that includes a light source configured to It can work in at least two modes, the at least two modes including a first mode and a second mode, and wherein, in the first mode, the light emitted by the light source is controlled by a light source control signal having a first frequency The attribute of the light source continuously changes at the first frequency, so that stripes appear on the image of the light source obtained when the light source is photographed by the rolling shutter camera, and in the second mode, the light source When the shutter camera shoots the light source The image of the light source does not exhibit stripes or exhibits stripes that are different from the stripes in the first mode.

Preferably, the first mode and the second mode are used to convey different information.

Preferably, in the second mode, the attribute of the light emitted by the light source is controlled to continuously change at the second frequency by a light source control signal having a second frequency different from the first frequency, so as to continuously change through the scroll The image of the light source obtained when the shutter camera shoots the light source does not present stripes or presents stripes different from the stripes in the first mode.

Preferably, the second frequency is greater than the first frequency.

Preferably, in the second mode, the attribute of the light emitted by the light source continuously changes at the first frequency, and the image of the light source obtained when the light source is photographed by the rolling shutter camera The image shows stripes different from the stripes in the first mode.

Another aspect of the present invention relates to a method for guiding a machine capable of autonomous movement using the above-mentioned system, which includes: transmitting information to a surrounding optical communication device through a rolling shutter camera installed on the machine capable of autonomous movement Collect and identify the transmitted information; determine whether the optical communication device is a target optical communication device based on the transmitted information; and if the optical communication device is a target optical communication device, control the autonomously movable machine Or part of it travels toward the optical communication device.

Preferably, the above method further includes: if the optical communication device is not a target optical communication device, identifying the optical communication device based on the information transmitted by the optical communication device, and obtaining the difference between the optical communication device and the target optical communication device The relative position relationship between; Determine the relative positional relationship between the autonomously movable machine or its part and the optical communication device; determine the relative positional relationship between the target optical communication device and the autonomously movable machine or its part; and at least The autonomously movable machine or its part is guided to the target optical communication device based in part on the relative positional relationship between the target optical communication device and the autonomously movable machine or its part.

Preferably, wherein determining the relative positional relationship between the autonomously movable machine or its part and the optical communication device includes: determining the relationship between the autonomously movable machine or its part and the optical communication device through relative positioning. The relative position relationship between.

Preferably, wherein, collecting and recognizing the information transmitted by a surrounding optical communication device through a rolling shutter camera installed on the machine capable of moving autonomously includes: obtaining the light through the rolling shutter camera. Continuous multi-frame images of the communication device; for each frame of image, determine whether there are stripes on the portion of the image corresponding to the position of the light source or what type of stripes exist; and determine each frame of image The information represented.

Preferably, the above method further includes: first controlling the autonomously movable machine to travel near the target optical communication device.

Preferably, wherein, first controlling the autonomously movable machine to travel near the target optical communication device includes: guiding the autonomously movable machine to the vicinity of the target optical communication device at least in part through a satellite navigation system; and/ Or, at least partially utilize the relative positional relationship between the other optical communication device and the target optical communication device to guide the autonomously movable machine to the vicinity of the target optical communication device.

Preferably, wherein, at least partially using the relative position relationship between the other optical communication device and the target optical communication device to guide the autonomously movable machine to the vicinity of the target optical communication device includes: the autonomously movable device The machine recognizes other optical communication devices while traveling, and obtains the relative positional relationship between the other optical communication device and the target optical communication device; determines the relative position between the autonomously movable machine and the other optical communication device Positional relationship; determining the relative positional relationship between the target optical communication device and the autonomously movable machine; and based at least in part on the relative positional relationship between the target optical communication device and the autonomously movable machine The machine capable of autonomous movement is guided to the vicinity of the target optical communication device.

Preferably, wherein, determining whether the optical communication device is a target optical communication device based on the transmitted information includes: determining whether the transmitted information explicitly or implicitly includes predetermined information.

Preferably, wherein the predetermined information is a predetermined identification word or verification code.

Preferably, wherein, determining whether the optical communication device is a target optical communication device based on the transmitted information includes: judging by the machine capable of autonomous movement whether the optical communication device is a target optical communication device; or, the capable The autonomously moving machine sends the delivered information to the server, and the server determines whether the optical communication device is a target optical communication device based on the delivered information, and sends the determination result to the autonomously movable machine .

Another aspect of the present invention relates to a machine capable of autonomous movement, including a rolling shutter camera, a processor, and a memory. The memory stores a computer program that can be used when executed by the processor Implement the above method.

Another aspect of the present invention relates to a storage medium in which a computer program is stored, and the computer program can be used to implement the above method when executed.

The embodiments of the present invention will be further described below with reference to the accompanying drawings, in which: FIG. 1 is a schematic diagram of a CMOS imaging device; FIG. 2 is a direction diagram of the CMOS imaging device to acquire images; FIG. 3 is a light source according to an embodiment of the present invention; 4 is a light source according to another embodiment of the present invention; FIG. 5 is an imaging timing diagram of a CMOS imaging device; FIG. 6 is another imaging timing diagram of a CMOS imaging device; FIG. 7 shows when the light source works in the first mode Fig. 8 shows the imaging timing chart of the CMOS imaging device when the light source is working in the first mode according to an embodiment of the present invention; Fig. 9 shows the imaging diagram of the CMOS imaging device at different stages; An imaging timing diagram of the CMOS imaging device when the light source is working in the second mode in one embodiment; FIG. 10 shows an imaging timing diagram of the CMOS imaging device when the light source is working in the first mode according to another embodiment of the present invention; FIG. 11 shows an imaging timing diagram of a CMOS imaging device for implementing different fringes from FIG. 8 according to another embodiment of the present invention; FIGS. 12-13 show two types of light sources obtained under different settings. Fringe image; Figure 14 shows a fringe-free image of the obtained light source; Figure 15 is an imaging diagram of a light label using three independent light sources according to an embodiment of the present invention; Figure 16 is an image according to the present invention An imaging diagram of an optical label including a positioning mark according to an embodiment; FIG. 17 shows an optical label including a reference light source and two data light sources according to an embodiment of the present invention; An imaging timing diagram of the CMOS imaging device of the optical tag is shown; and FIG. 19 shows a method for guiding the drone through the optical tag according to an embodiment of the present invention.

In order to make the objectives, technical solutions, and advantages of the present invention clearer, the following further describes the present invention in detail through specific embodiments with reference to the accompanying drawings.

An embodiment of the present invention relates to an optical communication device that can transmit different information by emitting different lights. The optical communication device is also referred to as an "optical tag" in this text, and the two can be used interchangeably throughout this application. The optical communication device includes a light source and a controller, and the controller can control the light source to operate in two or more modes through a light source control signal, the two or more modes including a first mode and a second mode, Wherein, in the first mode, the light source control signal has a first frequency, so that the properties of the light emitted by the light source continuously change at the first frequency to transmit The first information, in the second mode, the attribute of the light emitted by the light source changes continuously or does not change at the second frequency, so as to transmit the second information different from the first information.

The attribute of light in this application refers to any attribute that can be recognized by a CMOS imaging device. For example, it can be an attribute that is perceivable by the human eye such as the intensity, color, and wavelength of light, or it can be other attributes that are not perceivable by the human eye, such as Changes in the intensity, color, or wavelength of electromagnetic wavelengths outside the visible range of the human eye, or any combination of the above attributes. Therefore, the attribute change of light can be a single attribute change, or a combination of two or more attributes. When choosing the intensity of light as the attribute, it can be achieved simply by choosing to turn on or off the light source. In the following, for the sake of simplicity, the light source is turned on or off to change the properties of light, but those skilled in the art will understand that other methods for changing the properties of light are also feasible. It should be noted that the attribute of the light that changes at the first frequency in the first mode may be the same or different from the attribute of the light that changes at the second frequency in the second mode. Preferably, the properties of the light changed in the first mode and the second mode are the same.

When the light source works in the first mode or the second mode, a rolling shutter imaging device (such as a CMOS imaging device or a device with a CMOS imaging device (such as a mobile phone, a tablet computer, smart glasses, etc.)) can be used to image the light source, that is , Perform imaging by rolling shutter. In the following, a mobile phone is used as an example of a CMOS imaging device, as shown in Figure 2. The line scan direction of the mobile phone is shown as the vertical direction in FIG. 2, but those skilled in the art can understand that the line scan direction may also be the horizontal direction depending on the setting of the underlying hardware.

The light source can be a light source of various forms, as long as one of its properties that can be sensed by the CMOS imaging device can be changed at different frequencies. For example, the light source can be an LED light, an array composed of multiple LED lights, a display screen or a part of it, or even light irradiation Areas (such as areas where light is irradiated on a wall) can also be used as light sources. The shape of the light source can be various shapes, such as circular, square, rectangular, striped, L-shaped, cross-shaped, spherical, and the like. The light source can include various common optical devices, such as light guide plates, diffusers, diffusers, and so on. In a preferred embodiment, the light source may be a two-dimensional array composed of a plurality of LED lights, and one dimension of the two-dimensional array is longer than the other. Preferably, the ratio between the two is about 6-12:1. For example, the LED lamp array may be composed of a plurality of LED lamps arranged in a row. When emitting light, the LED lamp array can appear as a roughly rectangular light source, and the controller controls the operation of the light source.

Fig. 3 shows a light source according to an embodiment of the invention. When using a CMOS imaging device to image the light source shown in FIG. 3, it is preferable to make the long side of the light source shown in FIG. 3 perpendicular to the row direction of the CMOS imaging device (for example, the row scanning direction of the mobile phone shown in FIG. 2) Or roughly vertical to image as many stripes as possible under the same other conditions. However, sometimes users do not know the line scanning direction of their mobile phones. In order to ensure that the mobile phone can be recognized in various postures, and the maximum recognition distance can be achieved in both vertical and horizontal screens, the light source can be multiple rectangular Combine, for example, an L-shaped light source as shown in FIG. 4.

In another embodiment, the light source may not be limited to a planar light source, but may be implemented as a three-dimensional light source, for example, a bar-shaped cylindrical light source, a cube light source, and so on. For example, the light source can be placed on a square or suspended in the approximate center of an indoor place (such as a restaurant, a meeting room, etc.), so that nearby users in various directions can shoot the light source with a mobile phone to obtain the transmission of the light source. Information.

FIG. 5 shows an imaging timing chart of the CMOS imaging device, where each row corresponds to a row of sensors of the CMOS imaging device. Perform imaging in each row of the CMOS imaging sensor array When, it mainly involves two stages, namely exposure time and readout time. The exposure time of each line may overlap, but the readout time will not overlap.

It should be noted that FIG. 5 only schematically shows a small number of lines. In actual CMOS imaging devices, depending on the difference in resolution, there are usually thousands of lines of sensors. For example, for 1080p resolution, it has 1920×1080 pixels, the digit 1080 means that there are 1080 scanning lines, and the digit 1920 means that each line has 1920 pixels. For 1080p resolution, the readout time for each line is approximately 8.7 microseconds (ie, 8.7×10 ^-6 seconds).

If the exposure time is too long and the exposure time between adjacent rows overlaps a lot, it may show obvious transitional stripes during imaging. For example, multiple stripes between pure black pixel rows and pure white pixel rows have different grays. Pixel rows in degrees. The present invention expects to be able to present as clear pixel rows as possible. For this reason, the exposure time of CMOS imaging devices (such as mobile phones) can be set or adjusted (for example, through the APP installed on the mobile phone) to select the relative Short exposure time. In a preferred embodiment, the exposure time can be made approximately equal to or less than the readout time of each row. Taking 1080p resolution as an example, the readout time of each line is approximately 8.7 microseconds. In this case, you can consider adjusting the exposure time of the mobile phone to approximately 8.7 microseconds or less. FIG. 6 shows an imaging timing chart of the CMOS imaging device in this case. In this case, the exposure time of each row basically does not overlap, or the overlap portion is small, so that stripes with relatively clear boundaries can be obtained during imaging, which can be more easily recognized. It should be noted that FIG. 6 is only a preferred embodiment of the present invention, and longer (for example, equal to or less than twice, three times, or four times the readout time of each row, etc.) or shorter exposure time is also feasible of. For example, in the imaging process of the striped image shown in FIGS. 12 and 13 of the present application, the readout time of each line is approximately 8.7 microseconds, and the set exposure time of each line is 14 microseconds. In addition, in order to show stripes, a period of the light source can be The duration is set to about twice the exposure duration or longer, and preferably can be set to about four times the exposure duration or longer.

Figure 7 shows the imaging diagrams on the CMOS imaging device at different stages when the controller is used to make the light source work in the first mode. In the first mode, the properties of the light emitted by the light source are changed at a certain frequency. In this example Middle is to turn on and off the light source.

The upper part of FIG. 7 shows the state change diagrams of the light source at different stages, and the lower part shows the imaging diagrams of the light source on the CMOS imaging device at different stages, where the row direction of the CMOS imaging device is the vertical direction, and from the left Scan to the right. Since the CMOS imaging device collects images line by line, when shooting high-frequency flicker signals, the part of the obtained frame of image corresponding to the imaging position of the light source will form stripes as shown in the lower part of Figure 7. Ground, in time period 1, the light source is turned on, and the leftmost part of the scan line exposed in this time period shows bright stripes; in time period 2, the light source is turned off, and the scan line exposed during this time period shows dark stripes; in time period 3, the light source is turned on , The scan lines exposed in this period show bright stripes; in period 4, the light source is turned off, and the scan lines exposed in this period show dark stripes.

The light source control signal can be used to set the frequency of light source flicker, or to set the duration of each turn on and turn off of the light source to adjust the width of the stripes that appear. A longer turn-on or turn-off time usually corresponds to a wider stripe. For example, for the situation shown in Figure 6, if the duration of each turn on and turn off of the light source is set to be approximately equal to the exposure time of each line of the CMOS imaging device (the exposure time can be set through the APP installed on the mobile phone or manually Setting), the stripes with a width of only one pixel can be presented during imaging. In order to realize the long-distance recognition of the optical label, the stripes should be as narrow as possible. However, in practice, due to light interference, synchronization, etc., stripes with a width of only one pixel may be unstable or difficult to recognize. Therefore, in order to improve the stability of recognition, Preferably, stripes with a width of two pixels are implemented. For example, for the situation shown in FIG. 6, it is possible to achieve a width of about two paintings by setting the duration of each turn on or turn off of the light source approximately equal to about 2 times the exposure duration of each row of the CMOS imaging device. The pixel stripes are specifically shown in Fig. 8, where the signal in the upper part of Fig. 8 is a light source control signal, and its high level corresponds to turning on the light source, and low level corresponds to turning off the light source. In the embodiment shown in FIG. 8, the duty cycle of the light source control signal is set to about 50%, and the exposure time of each row is set to be approximately equal to the readout time of each row, but those skilled in the art can understand that, Other settings are also feasible, as long as they can show distinguishable fringes. For simplicity of description, the synchronization between the light source and the CMOS imaging device is used in FIG. 8 so that the turn-on and turn-off time of the light source roughly corresponds to the start or end time of the exposure duration of a certain line of the CMOS imaging device, but Those skilled in the art can understand that even if the two are not synchronized as shown in Figure 8, there may be obvious stripes on the CMOS imaging device. At this time, there may be some transitional stripes, but there must be some exposure when the light source is always turned off. The line (that is, the darkest stripe) and the line that is exposed when the light source is always on (that is, the brightest stripe) are separated by one pixel. The light and dark changes (that is, fringes) of such pixel rows can be easily detected (for example, by comparing the brightness or grayscale of some pixels in the imaging area of the light source). Furthermore, even if there are no lines that are exposed when the light source is always off (that is, the darkest stripes) and lines that are exposed when the light source is always on (that is, the brightest stripes), if there is a light source that is turned on during the exposure time, t1 is less than a certain time The length or the line with a smaller proportion of the entire exposure time (ie darker stripes), and the line with the light source turn-on part t2 greater than a certain length of time or a larger percentage of the entire exposure time (ie brighter stripes) during the exposure time , And t2-t1>light-dark fringe difference threshold (for example, 10 microseconds), or t2/t1>light-dark fringe ratio threshold (for example, 2), the light-dark changes between these pixel rows can also be detected. The above-mentioned light and dark fringe difference threshold and ratio threshold are related to the luminous intensity of the light label, the properties of the photosensitive device, and the shooting distance. Those skilled in the art can understand, Other thresholds are also feasible, as long as they can show fringes that can be distinguished by a computer. When the fringe is recognized, the information transmitted by the light source at this time can be determined, such as binary data 0 or data 1.

The fringe recognition method according to an embodiment of the present invention is as follows: Obtain the image of the light label, and use projection to segment the imaging area of the light source; collect data in different configurations (for example, different distances, different light source flicker frequencies, etc.) Striped pictures and non-striped pictures; normalize all collected pictures to a specific size, such as 64*16 pixels; extract each pixel feature as an input feature to build a machine learning classifier; perform two-class discrimination to judge Is it a striped picture or a non-striped picture. For fringe recognition, those of ordinary skill in the art can also use any other methods known in the art for processing, which will not be described in detail.

For a strip light source with a length of 5 cm, when using a common mobile phone on the market and setting the resolution to 1080p, when shooting at a distance of 10 meters (that is, the distance is 200 times the length of the light source), The strip light source occupies about 6 pixels in its length direction. If the width of each stripe is 2 pixels, there will be multiple distinct pixels within the width range of the 6 pixels. At least one obvious streak will appear, which can be easily identified. If you set a higher resolution or use an optical zoom, the fringes can also be recognized at a longer distance, such as 300 or 400 times the length of the light source.

The controller can also make the light source work in the second mode. In one embodiment, in the second mode, the light source control signal may have another frequency different from the first mode to change the properties of the light emitted by the light source, such as turning the light source on and off. In one embodiment, compared to the first mode, the controller can increase the frequency of turning on and turning off the light source in the second mode. For example, the frequency of the first mode may be greater than or equal to 8000 times/sec, and the frequency of the second mode may be greater than the frequency of the first mode. For the situation shown in Figure 6, the light source can be configured to expose each row of the CMOS imaging device. The light source is turned on and off at least once during the light time. Figure 9 shows a situation where the light source is turned on and off only once during the exposure time of each line. The signal in the upper part of Figure 9 is the light source control signal, and its high level corresponds to the light source turning on, and the low level corresponds to Turn off the light source. Since the light source will be turned on and off in the same way during the exposure time of each line, the exposure intensity obtained at each exposure time is roughly the same, so there will be no significant difference in the brightness of each pixel row of the final image of the light source , So there are no stripes. Those skilled in the art can understand that higher opening and closing frequencies are also feasible. In addition, for simplicity of description, the synchronization between the light source and the CMOS imaging device is used in FIG. 9 so that the turn-on time of the light source roughly corresponds to the start time of the exposure duration of a certain line of the CMOS imaging device. The person can understand that even if the two are not synchronized as shown in FIG. 9, there will be no significant difference in the brightness of each pixel row in the final imaging of the light source, so there is no fringe. When the fringe cannot be identified, the information transmitted by the light source at this time can be determined, such as binary data 1 or data 0. For the human eye, the light source of the present invention does not perceive any flicker when it works in the first or second mode. In addition, in order to avoid the flicker that may be perceived by the human eye when switching between the first mode and the second mode, the duty cycle of the first mode and the second mode can be set to be approximately equal, so as to achieve different modes The luminous flux is roughly the same.

In another embodiment, in the second mode, direct current can be supplied to the light source, so that the light source emits light whose properties do not substantially change, so that the light source obtained when the light source is photographed by the CMOS image sensor There are no stripes on one frame of the light source. In addition, in this case, it is also possible to achieve substantially the same luminous flux in different modes, so as to avoid flicker that may be perceived by the human eye when switching between the first mode and the second mode.

Figure 8 above describes an embodiment in which stripes are presented by changing the intensity of the light emitted by the light source (for example, by turning on or off the light source). In another embodiment, as shown in Figure 10 As shown, the fringes can also be presented by changing the wavelength or color of the light emitted by the light source. In the embodiment shown in FIG. 10, the light source includes a red lamp that emits red light and a blue lamp that emits blue light. The two signals in the upper part of FIG. 10 are respectively a red light control signal and a blue light control signal, wherein the high level corresponds to turning on the corresponding light source, and the low level corresponds to turning off the corresponding light source. The phases of the red light control signal and the blue light control signal are offset by 180°, that is, the levels of the two are opposite. Through the red light control signal and the blue light control signal, the light source can alternately emit red light and blue light outwards, so that red and blue stripes can be presented when the CMOS imaging device is used to image the light source.

By determining whether there are fringes in a part of the image taken by the CMOS imaging device corresponding to the light source, the information conveyed by each frame of the image, such as binary data 1 or data 0, can be determined. Further, by capturing continuous multiple frames of images of the light source through the CMOS imaging device, an information sequence composed of binary data 1 and 0 can be determined, and information transmission from the light source to the CMOS imaging device (such as a mobile phone) can be realized. In one embodiment, when the CMOS imaging device is used to capture continuous multi-frame images of the light source, it can be controlled by the controller so that the switching time interval between the working modes of the light source is equal to the imaging time of a complete frame of the CMOS imaging device Length to achieve frame synchronization between the light source and the imaging device, that is, 1 bit of information is transmitted per frame. For a shooting speed of 30 frames per second, 30 bits of information can be transmitted per second, and the encoding space can reach 2 ^30. The information can include, for example, the start frame mark (frame header), the ID of the light tag, the password, and the verification code. , URL information, address information, timestamp or different combinations thereof, etc. The sequence relationship of the above-mentioned various information can be set according to a structured method to form a data packet structure. Each time a complete data packet structure is received, it is regarded as obtaining a set of complete data (a data packet), which can then be read and verified and analyzed. The following table shows the data packet structure according to an embodiment of the present invention:

In the above description, the information conveyed by the frame of image is determined by judging whether there are stripes at the imaging position of the light source in each frame of image. In other embodiments, the different information conveyed by the frame of image can be determined by identifying different stripes at the imaging position of the light source in each frame of image. For example, in the first mode, the property of the light emitted by the light source is controlled to continuously change at the first frequency through the light source control signal with the first frequency, so that the light source can be obtained when the light source is photographed by the CMOS image sensor The first stripe appears on the image; in the second mode, the light source control signal with the second frequency is used to control the properties of the light emitted by the light source to continuously change at the second frequency, so that it can pass through the CMOS image sensor The image of the light source obtained when the light source is photographed shows a second fringe different from the first fringe. The difference in stripes may be based on, for example, different widths, colors, brightness, etc., or any combination thereof, as long as the difference can be recognized.

In one embodiment, stripes of different widths can be realized based on different light source control signal frequencies. For example, in the first mode, the light source can work as shown in FIG. 8 to achieve a width of approximately two pixels. The first type of stripe; in the second mode, the duration of the high level and low level in each cycle of the light source control signal in Figure 8 can be modified to twice the original, as shown in Figure 11. Thus, the second type of stripes with a width of about four pixels is realized.

In another embodiment, stripes of different colors can be realized. For example, the light source can be set to include a red light emitting red light and a blue light emitting blue light. In the first mode, the blue light can be turned off , And make the red light work as shown in Figure 8 to achieve red and black stripes; in the second mode, you can turn off the red light and make the blue light work as shown in Figure 8 to achieve blue and black stripes . In the above-mentioned embodiment, in the first mode and the second mode, the same The frequency light source control signal realizes red and black stripes and blue and black stripes, but it is understood that light source control signals with different frequencies can also be used in the first mode and the second mode.

In addition, those skilled in the art can understand that more than two kinds of stripes can be implemented to represent more than two kinds of information. For example, in the above embodiment where the light source includes a red light and a blue light, a third mode can be further set. In this third mode, the red and blue lights are controlled in the manner shown in FIG. 10 to achieve red and blue stripes, that is, the third information. Obviously, optionally, another type of information, that is, the fourth type of information, can be further transmitted through a fourth mode that does not show stripes. You can arbitrarily select multiple of the above four modes for information transmission, or further combine with other modes, as long as different modes produce different stripe patterns.

Figure 12 shows that for LED lights that flicker at a frequency of 16000 times per second (the duration of each cycle is 62.5 microseconds, where the on-duration and off-duration are each about 31.25 microseconds), using 1080p resolution The stripe on the image obtained through experiments when the exposure time of each line is set to 14 microseconds. It can be seen from Fig. 12 that stripes with a width of approximately 2-3 pixels are present. Figure 13 shows that after adjusting the flashing frequency of the LED lights in Figure 12 to 8000 times per second (the duration of each cycle is 125 microseconds, where the on-duration and off-duration are each about 62.5 microseconds), after Streaks on the image obtained through experiments under other conditions unchanged. It can be seen from Figure 13 that stripes with a width of approximately 5-6 pixels are present. Figure 14 shows that after adjusting the blinking frequency of the LED lights in Figure 12 to 64000 times per second (the duration of each cycle is 15.6 microseconds, where the on-duration and off-duration are each about 7.8 microseconds), after There are no stripes on the image obtained through experiments under other conditions unchanged. The reason is that the 14 microseconds of exposure time per line basically covers an on time and an off time of the LED light.

In the foregoing, for convenience of description, a square wave is used as an example to describe a light source control signal with a corresponding frequency, but those skilled in the art will understand that the light source control signal can also use other waveforms, such as sine waves, triangular waves, and the like.

The foregoing describes the case where one light source is used. In some embodiments, two or more light sources may also be used. The controller can independently control the operation of each light source. 15 is an imaging diagram of a light label using three independent light sources according to an embodiment of the present invention, in which stripes appear in the imaging position of two light sources, and no stripes appear in the imaging position of one light source. One frame of image can be used to convey information, such as binary data 110.

In one embodiment, the optical label may also include one or more positioning marks located near the information transmission light source. The positioning marks may be, for example, lights of a specific shape or color, and the lights may be kept on during work, for example. The location mark can help users of CMOS imaging devices (such as mobile phones) to easily find the optical tag. In addition, when the CMOS imaging device is set to a mode of photographing the light tag, the imaging of the positioning mark is more obvious and easy to identify. Therefore, one or more positioning marks arranged near the information transmission light source can also help the mobile phone to quickly determine the position of the information transmission light source, thereby helping to identify whether there are stripes in the imaging area corresponding to the information transmission light source. In one embodiment, when identifying whether there are stripes, the positioning mark can be identified in the image first, so that the approximate position of the light label can be found in the image. After the positioning mark is identified, one or more regions in the image can be determined based on the relative position relationship between the positioning mark and the information transmission light source, and the area covers the imaging position of the information transmission light source. Then, these regions can be identified to determine whether there are stripes or what kind of stripes exist. Fig. 16 is an imaging diagram of a light tag including a positioning mark according to an embodiment of the present invention, which includes three horizontally arranged information transmission light sources, and two vertical positioning mark lights located on both sides of the information transmission light source .

In one embodiment, the optical tag may include an ambient light detection circuit, and the ambient light detection circuit may be used to detect the intensity of the ambient light. The controller can adjust the intensity of the light emitted by the light source when it is turned on based on the intensity of the detected ambient light. For example, when the ambient light is relatively strong (for example, during the day), the intensity of the light emitted by the light source is relatively large, and when the ambient light is relatively weak (for example, at night), the intensity of the light emitted by the light source is relatively small.

In one embodiment, the optical tag may include an ambient light detection circuit, and the ambient light detection circuit may be used to detect the frequency of the ambient light. The controller may adjust the frequency of light emitted by the light source when the light source is turned on based on the frequency of the detected ambient light. For example, when the ambient light has the same stroboscopic light source, switch the light emitted by the light source to another unoccupied frequency.

In the actual application environment, if there is a lot of noise, or when the recognition distance is very far, it may affect the accuracy of recognition. Therefore, in order to improve the accuracy of recognition, in one embodiment of the present invention, in addition to the above-mentioned light source for transmitting information (for clarity, it will be referred to as "data light source" hereinafter) in the optical label. , Can also include at least one reference light source. The reference light source itself is not used to transmit information, but is used to assist in identifying the information transmitted by the data light source. The reference light source can be similar in physical structure to the data light source, but works in a predetermined working mode, which can be one or more of various working modes of the data light source. In this way, the decoding of the data light source can be converted into the calculation of matching (for example: correlation) with the image of the reference light source, thereby improving the accuracy of decoding.

Figure 17 shows a light label including a reference light source and two data light sources according to an embodiment of the present invention, wherein three light sources are arranged side by side, the first light source is used as the reference light source, and the other two light sources are respectively used as the first light source. A data light source and a second data light source. It should be noted that the number of reference light sources in the light label can be one or more, not limited to one; similarly, the number of reference light sources The number of material light sources can also be one or more, and is not limited to two. In addition, because the reference light source is used to provide auxiliary identification, its shape and size are not necessarily the same as the data light source. For example, in one embodiment, the length of the reference light source may be half of the data light source.

In one embodiment, each of the first data light source and the second data light source shown in FIG. 17 is configured to work in three modes, for example, to display a fringe-free image with a stripe width of 2 pixels. An image with a stripe width of 4 pixels. The reference light source can be configured to always work in one of the three modes to display one of the above three images, or alternately work in different modes to alternately display any two or all of the above three images in different frames, thereby Provide a benchmark or reference for the image recognition of the data light source. Taking the reference light source alternately displaying images with a stripe width of 2 pixels and an image with a stripe width of 4 pixels in different frames as an example, the image of the data light source in each frame can be the same as the current frame and one The image of the reference light source in the adjacent frame (such as the previous frame or the following frame) (these images must contain an image with a stripe width of 2 pixels and an image with a stripe width of 4 pixels). Compare to determine the type of the image; alternatively, it is also possible to collect consecutive multiple frames of images of the reference light source within a period of time, and use the odd-numbered image and the even-numbered image as a group, for each The characteristics of the group of images are averaged (for example, the average value of the stripe width of each group of images), and according to the stripe width to distinguish which group of images corresponds to an image with a stripe width of 2 pixels or a stripe width For an image of 4 pixels, the average feature of an image with a stripe width of 2 pixels and an average feature of an image with a stripe width of 4 pixels can be obtained. After that, it can be determined that the data light source is in each frame Does the image meet one of these average characteristics.

Since the reference light source and the data light source are located at approximately the same position and are subject to the same ambient lighting conditions, interference, noise, etc., it can provide one or more reference images or reference images for image recognition in real time, thereby Can improve the recognition of the information transmitted by the data light source Accuracy and stability. For example, by comparing the imaging of the data light source with the imaging of the reference light source, the working mode of the data light source can be accurately identified, thereby identifying the data transmitted by it.

Furthermore, according to the CMOS imaging principle, when multiple light sources change their properties at the same frequency but different phases, fringe patterns of the same width but different phases will be produced. The fringe patterns of the same width but different phases can be matched using a matching method. Determine accurately. In one embodiment, the reference light source can be controlled to work in a predetermined working mode. In this working mode, the image of the reference light source may show stripes with a width of 4 pixels, for example. At this time, if the data light source is controlled to work in this working mode at the same time, and the phase of the data light source and the reference light source are consistent, the fringes appearing on the image of the data light source are similar to the fringes appearing on the image of the reference light source (For example, the width is also 4 pixels) and there is no phase difference; if the data light source is controlled to work in this working mode at the same time, but the phase of the data light source and the reference light source are inconsistent (for example, antiphase or 180° phase difference), then The fringes appearing on the image of the reference light source are similar to the fringes appearing on the image of the reference light source (for example, the width is also 4 pixels) but there is a phase difference.

FIG. 18 shows an imaging timing chart of the CMOS imaging device for the optical label shown in FIG. 17. The upper part of FIG. 18 shows the respective control signals of the reference light source, the first data light source, and the second data light source. The high level may correspond to the turning on of the light source, and the low level may correspond to the turning off of the light source. As shown in FIG. 18, the frequencies of the three control signals are the same, and the phases of the first data light source control signal and the reference light source control signal are consistent, and the phases of the second data light source control signal and the reference light source control signal are 180° out of phase. In this way, when the CMOS imaging device is used to image the light tag, the imaging of the reference light source, the first data light source and the second data light source will all show stripes with a width of approximately 4 pixels, but the first data light source It is consistent with the phase of the fringe on the imaging of the reference light source (for example, the line where the bright fringe of the reference light source is located and the line where the bright fringe of the first data light source is located Is consistent, the row of the dark fringe of the reference light source is the same as the row of the dark fringe of the first data light source, and the phase of the fringe on the imaging of the second data light source and the reference light source is reversed (for example, reference The row where the bright stripes of the light source are located is the same as the row where the dark stripes of the second data light source are located, and the row where the dark stripes of the reference light source are located is the same as the row where the bright stripes of the second data light source are located).

By providing a reference light source and adopting phase control for the data light source, the amount of information that the data light source can transmit each time can be further improved while improving the recognition ability. For the optical label shown in FIG. 17, if the first data light source and the second data light source are configured to work in the first mode and the second mode, there are no stripes in the first mode, but in the second mode. stripe. If no reference light source is provided, each data light source can transmit one of two kinds of data in a frame of image, such as 0 or 1. By providing the reference light source and making it work in the second mode, and further providing phase control when the data light source works in the second mode, the second mode itself can be used to transmit more than one type of data. Taking the method shown in FIG. 18 as an example, the second mode combined with phase control can itself be used to transmit one of two kinds of data, so that each data light source can transmit one of three kinds of data in one frame of image.

In the above manner, by introducing a reference light source, the phase control of the data light source can be realized, so the encoding density of the data light source of the optical label can be increased, and the encoding density of the entire optical label can be increased accordingly. For example, for the embodiment described above, if the reference light source is not used (that is, the reference light source is used as the third data light source), each data light source can transmit one of two kinds of data in one frame of image, so the entire The light tag (including three data light sources) can transmit one of 2 or ³ data combinations in one frame of image; and if a reference light source is used, each data light source can transmit one of three types of data in one frame of image, so the entire light tag (comprising two data sources) can pass one of three ^two kinds of data combined in one image. If the number of data light sources in the light label is increased, the effect will be more obvious. For example, for the embodiment described above, if a light label containing five light sources is used, the entire light label ( Contains five data light sources) One of ²⁵ data combinations can be transferred in one frame of image; and when one of the light sources is selected as the reference light source, the entire light label (including four data light sources) is one frame of image One of 3 or ⁴ data combinations can be passed in. Similarly, by increasing the number of reference light sources in the optical label, the encoding density of the entire optical label can also be further improved. The following provides some experimental data for matching the image of the data light source and the image of the reference light source for matching calculation (for example: correlation calculation). Among them, the meaning of the calculation result is defined as follows: 0.00~±0.30 Micro correlation ±0.30~±0.50 Real correlation ±0.50~±0.80 Significant correlation ±0.80~±1.00 High correlation

Among them, a positive value represents a positive correlation, and a negative value represents a negative correlation. If the frequency and phase of the data light source and the reference light source are the same, then in an ideal state, the images of the two light sources are exactly the same, and the result of the correlation calculation is +1, indicating a complete positive correlation. If the frequency of the data light source and the reference light source are the same, but the phases are opposite, then in an ideal state, the image fringe width of the two light sources is the same, but the positions of the bright and dark fringes are exactly opposite, so the correlation calculation result is − 1, indicates a complete negative correlation. It can be understood that in the actual imaging process, due to interference, errors, etc., it is difficult to obtain images with a completely positive correlation and a completely negative correlation. If the data light source and the reference light source work in different working modes to display stripes of different widths, or one of them does not display stripes, the images of the two are usually slightly correlated.

The following Table 1 and Table 2 respectively show the correlation calculation results when the data light source and the reference light source adopt the same frequency and the same phase, and the correlation calculation results when the data light source and the reference light source adopt the same frequency and opposite phase. For each situation, five images were taken, and the correlation calculation between the reference light source image in each frame image and the data light source image in the frame image was performed.

It can be seen from the above table that when the data light source and the reference light source use the same frequency and the same phase, the correlation calculation results can show that they are significantly positively correlated. When the data source and parameters When considering the light source with the same frequency and opposite phase, the correlation calculation result can show that the two are significantly negatively correlated.

Compared with the recognition distance of about 15 times of the two-dimensional code in the prior art, the recognition distance of at least 200 times of the optical label of the present invention has obvious advantages. This long-distance recognition capability is especially suitable for outdoor recognition. Taking a recognition distance of 200 times as an example, for a light source with a length of 50 cm installed on the street, people within 100 meters of the light source can communicate with the light source through a mobile phone. To interact. In addition, the solution of the present invention does not require the CMOS imaging device to be located at a fixed distance from the optical tag, nor does it require time synchronization between the CMOS imaging device and the optical tag, and does not require precise detection of the border and width of each stripe. Therefore, it has extremely strong stability and reliability in actual information transmission. In addition, the solution of the present invention does not require that the CMOS imaging device must be substantially facing the optical tag to perform identification, especially for optical tags with striped or spherical light sources. For example, for a strip or column optical label set on a square, all CMOS imaging devices within a 360° range around it can identify it. If the strip-shaped or column-shaped optical tag is placed on a wall, all CMOS imaging devices within approximately 180° around it can recognize it. For a spherical light label set on the square, any CMOS imaging device in the surrounding three-dimensional space can recognize it.

Due to the above-mentioned advantages of the optical tag of the present invention, it can be used to achieve precise guidance within the field of view, for example, to guide a machine that can move autonomously. An embodiment of the present invention relates to a system for guiding an autonomously movable machine through an optical label, which includes an autonomously movable machine and the optical label described in any of the above embodiments. A CMOS camera is installed on the machine that can move autonomously, which can collect and recognize the information transmitted by the optical tag.

The following describes the drone delivery application of online shopping as an example. Buyers can use their apartment as the delivery address and fill in the delivery address information in the online shopping platform, such as some of the following information: geographic location information, community information, building number, floor, etc. Buyers can place a light tag on the apartment (for example, the balcony, exterior wall, etc.) of the apartment to serve as the target light tag for the drone to deliver goods. After the buyer completes the shopping through the Internet, the optical tag can be configured to continuously work in different modes to deliver predetermined information. The predetermined information can be, for example, the ID information of the optical tag itself or the buyer’s online shopping platform. ID information, the verification code received by the buyer from the online shopping platform after shopping on the online shopping platform, etc., as long as the predetermined information is known to the online shopping platform and can be used to identify the buyer or the goods purchased. The online shopping platform can transmit the reservation information to the drone.

The method of the present invention for UAV guidance through optical tags may be as shown in Figure 19, which includes the following steps:

Step 101: Control the drone to travel near the target light tag.

After the drone picks up the goods to be sent to the buyer, it can first fly to the buyer's receiving address (that is, the buyer's apartment). In one embodiment, the delivery address may preferably be the geographic location information of the target light tag itself (for example, the precise latitude and longitude of the light tag, altitude information, etc.), and may also include other information, such as the target light tag The orientation information, etc.

Step 101 can be implemented in various possible existing ways in the art. For example, the drone can fly to the vicinity of the receiving address (that is, near the buyer's optical tag) through GPS navigation and other methods. Existing GPS navigation methods can reach an accuracy range of tens of meters, and the optical tag of the present invention can achieve at least 200 times the recognition distance. Taking 200 times the recognition distance as an example, for a light source of 20 cm in length, the drone only needs to be able to Fly to within 40 meters of the light source can be identified. In step In step 101, the relative position relationship between other optical tags and the target optical tags may also be used to guide the drone to the vicinity of the target optical tags. The relative positional relationship between the various optical tags can be stored in advance and can be obtained by the drone, for example. When flying, the drone can identify other optical tags along its flight path, and obtain the relative positional relationship between the other optical tags and the target optical tag. Then, the drone can use relative positioning (also called reverse positioning) To determine the relative position relationship between it and the other optical tags, so that the relative position relationship between the target optical tag and the drone can be determined. Based on the relative position relationship, the drone can be guided to the vicinity of the target light tag. Those skilled in the art can understand that it is also possible to use a combination of the above various methods to guide the drone to the vicinity of the target optical tag.

Various relative positioning methods known in the art can be used to determine the relative position relationship between the drone and the optical tag. In one embodiment, the drone can use its imaging device to perform image capture on the optical tag, and obtain its relative distance to the optical tag based on the captured image (the larger the image, the closer the distance; the smaller the image, the longer the distance Far), and the current orientation information of the drone can be obtained through the built-in sensor, and the relative direction of the drone and the light tag can be obtained based on the orientation information (preferably, the position of the light tag in the image can be further combined to make more Accurately determine the relative direction of the drone and the optical tag), so that the relative position relationship between the drone and the optical tag can be obtained based on the relative distance and relative direction between them. In addition, many imaging devices currently on the market are usually equipped with binocular cameras or depth cameras. The imaging devices equipped with binocular cameras or depth cameras are used to capture images of the optical tag, and the imaging device and light can be easily obtained. The relative distance between labels. In another embodiment, in order to determine the relative direction between the user and the optical tag, the orientation information of the optical tag can be stored in the server. After the user recognizes the identification information of the optical tag, the identification information can be obtained from the server The orientation information is then based on the orientation of the light tag The information and the perspective distortion of the imaging of the light tag on the user's mobile phone can be used to calculate the relative direction of the user and the light tag.

It should be noted that this step 101 is not a necessary step of the present invention, and it may be omitted in some cases. For example, if the target light tag itself is already in the field of view of the drone.

Step 102: Use the CMOS camera installed on the drone to collect information transmitted by a certain light tag around, and identify the transmitted information.

After flying near the buyer's optical tag, the drone can search for the optical tag in its field of view, and collect the information transmitted by the found optical tag through the CMOS camera installed on it and identify the transmitted information. For example, a drone can obtain a continuous multi-frame image of a light tag through its CMOS camera, and for each frame of image, determine whether there are stripes or what type of stripes exist in the part of the image corresponding to the position of the light source. Stripes, and determine the information represented by each frame of image. In one embodiment, if the drone finds a light tag in its field of view, but the information transmitted by it cannot be recognized due to the distance, the drone can approach the light tag appropriately to achieve the light tag Identification of the information transmitted.

Step 103: Determine whether the optical label is a target optical label based on the transmitted information.

The drone can determine whether the optical tag is a target optical tag based on the information transmitted by the optical tag. For example, the drone can determine whether the transmitted information explicitly or implicitly contains the predetermined information. If it does, it can be determined that the optical tag is a target optical tag; otherwise, it can be determined that the optical tag is not a target optical tag. In an embodiment, the drone itself can determine whether the optical tag is a target optical tag. In another embodiment, the drone can deliver the light tag The information is sent to a server that can communicate with the drone, and the server determines whether the optical tag is a target optical tag based on the transmitted information, and sends the judgment result to the drone. The information transmitted by the optical label can be encrypted information.

Step 104: If the optical tag is a target optical tag, control the drone to travel to the optical tag.

Since the drone has determined that a certain optical tag in the field of view is the flight destination, the drone can fly to the optical tag without error, for example, through the visual guidance of the optical tag. In one embodiment, the drone can use the existing ranging technology to stop at a certain distance from the optical tag, for example, a position several tens of centimeters away from the optical tag, to avoid collision with the optical tag. In one embodiment, the drone can perform relative positioning and adjust its flight route based on the perspective deformation of the image of the light tag captured by it, so that the drone can finally stop in a certain direction relative to the light tag, for example, Right in front of the light label. A shelf for receiving goods can be arranged directly in front of the light tag, and the drone can easily deliver the goods to the shelf.

If it is determined that the optical tag is not a target optical tag, the drone can identify other optical tags nearby, which is similar to the above process and will not be repeated. Preferably, in one embodiment, if it is determined that the optical tag is not the target optical tag, the drone can use the optical tag to determine the relative position relationship between it and the target optical tag. For example, the drone can determine the relative positional relationship between it and the optical tag through relative positioning, and can identify the optical tag based on the information transmitted by the optical tag (for example, obtain the identification information of the optical tag) and obtain the optical tag. The relative positional relationship between the tag and the target light tag (the relative positional relationship between each light tag can be stored in advance and can be obtained by the drone, for example), so that the relative position between the drone and the target light tag can be determined Positional relationship. After obtaining the relative position relationship, the drone can use the relative position relationship and optional other navigation information (for example, GPS information) to fly to the vicinity of the target light tag.

Those skilled in the art can understand that the drone delivery solution guided by optical tags of the present invention is not limited to optical tags arranged at the buyer’s apartment (for example, the balcony, exterior wall, etc.) of the apartment, and it is obviously also applicable For light tags arranged in more open areas, for example, light tags arranged in courtyards.

In addition, if the buyer does not have his own light tag, or wants to deliver the goods to the location where other light tags are located (for example, public light tags in squares, parks, etc., or light tags at a friend’s house), they can deliver the goods The relevant information (for example, ID information of the target light tag, geographic location information, etc.) of the optical tag at the delivery address (that is, the target optical tag) is notified to the online shopping platform. The online shopping platform can inform the drone of the corresponding information. After the drone flies near the target optical tag, it can identify the information transmitted by the nearby optical tag (for example, the ID information transmitted by the optical tag), and finally determine the target optical tag .

In addition, the drone delivery solution guided by optical tags of the present invention is not only applicable to optical tags with fixed positions, but also to non-fixed optical tags (for example, optical tags that can be carried by people). For example, if a buyer wishes to make online shopping during an event in the square and hopes to deliver the goods to his current location, he can inform the online shopping platform of his current geographic location information and turn on the light tag he carries with him. The optical tag can be configured to transmit predetermined information. The predetermined information can be, for example, the ID information of the optical tag itself, the ID information of the buyer on the online shopping platform, and the verification received from the platform after the buyer makes a purchase on the online shopping platform. Code, etc., as long as the predetermined information is known to the online shopping platform and can be used to identify the buyer or the goods purchased. After the drone flies near the buyer’s location, it can identify the nearby light tag delivery And finally determine the target light label (that is, the light label carried by the buyer) to complete the delivery of the goods. In one embodiment, the online shopping platform can inform the buyer of the estimated time of arrival of the drone, so that the buyer can freely move around during this period, as long as he returns to the vicinity of the previous location at the estimated time of arrival. In one embodiment, the buyer may not return to the previous location, but may send his new location to the online shopping platform, and the online shopping platform may notify the drone of the new location so that the drone can fly there. Near the new location. In one embodiment, the buyer can also set the delivery address of the goods to a certain address that is expected to arrive at a certain time, and instruct the online shopping platform to deliver the goods to the vicinity of the address at that time.

The above description takes the drone delivery application for online shopping as an example, but it is understandable that the drone guidance through optical tags is not limited to the above applications, but can be used for various applications that require precise positioning of the drone. For example, UAV automatic charging, UAV automatic parking, UAV line navigation and so on. In addition, those skilled in the art can understand that the optical tag-based guidance of the present invention is not only applicable to drones, but can also be applied to other types of autonomously moving machines, such as unmanned vehicles, robots, etc. CMOS cameras can be installed on unmanned cars or robots and can interact with light tags in a similar way to drones. In one embodiment, a part of the autonomously movable machine is movable, but another part is fixed. For example, the machine that can move autonomously may be a machine located on an assembly line or in a warehouse that usually has a fixed position. The main part of the machine can be fixed in most cases, but has one or more movable machines. arm. The CMOS camera can be installed in the fixed part of the machine to determine the position of the optical label, so that the movable part of the machine (for example, the robotic arm) can be guided to the position of the optical label. For this kind of machine, obviously, step 101 described above is not needed of. In addition, it can be understood that the CMOS camera can also be installed on a movable part of the machine, for example, on each robotic arm.

References to "various embodiments", "some embodiments", "one embodiment", or "an embodiment" herein refer to the specific features, structures, or properties described in connection with the embodiments included in In at least one embodiment. Therefore, the appearances of the phrases "in various embodiments", "in some embodiments", "in one embodiment", or "in an embodiment" in various places throughout this document do not necessarily refer to the same Examples. In addition, specific features, structures, or properties can be combined in any suitable manner in one or more embodiments. Therefore, a specific feature, structure, or property shown or described in one embodiment can be combined in whole or in part with the feature, structure, or property of one or more other embodiments without limitation, as long as the combination is not non-limiting. Logical or not working. Expressions similar to "according to A" or "based on A" appearing in this article mean non-exclusive, that is, "according to A" can cover "only according to A" or "according to A and B" unless specifically The statement or the context clearly knows its meaning as "only based on A". For clarity in this application, some illustrative operating steps are described in a certain order, but those skilled in the art can understand that each of these operating steps is not essential, and some of the steps can be omitted Or replaced by other steps. These operating steps do not have to be executed sequentially in the manner shown. On the contrary, some of these operating steps can be executed in different sequences according to actual needs, or executed concurrently, as long as the new execution method is not illogical or unable to work.

Thus, several aspects of at least one embodiment of the present invention have been described, and it can be understood that various changes, modifications and improvements can be easily made by those skilled in the art. Such changes, modifications and improvements are intended to be within the spirit and scope of the present invention.

Claims (18)

A system for guiding a machine that can move autonomously, including: a machine that can move autonomously, on which a rolling shutter camera is installed; and an optical communication device, which includes a light source configured to work in at least two types Mode, the at least two modes include a first mode and a second mode, and wherein, in the first mode, the properties of the light emitted by the light source are controlled by a light source control signal having a first frequency according to the first mode A frequency keeps changing so that stripes appear on the image of the light source obtained when the light source is photographed by the rolling shutter camera. In the second mode, when the rolling shutter camera is used for the image of the light source, Stripes are not present on the image of the light source obtained when the light source is photographed. The system according to claim 1, wherein, in the second mode, the light source control signal having the second frequency different from the first frequency controls the attribute of the light emitted by the light source to the second The frequency is continuously changed so that no stripes appear on the image of the light source obtained when the light source is photographed by the rolling shutter camera. The system according to claim 2, wherein the second frequency is greater than the first frequency. The system according to claim 1, wherein, in the second mode, the attribute of the light emitted by the light source continuously changes at the first frequency, and when the light source is photographed by the rolling shutter camera The obtained image of the light source presents stripes different from the stripes in the first mode. The system according to claim 1, wherein the light source is a bar light source or a spherical light source. The system according to claim 1, wherein the machine capable of autonomous movement includes a machine whose only part can move. A method for guiding a machine capable of autonomous movement using the system according to any one of claim items 1-6, comprising: using a rolling shutter camera installed on the machine capable of autonomous movement to detect a certain light in the surroundings. The information transmitted by the communication device is collected, and the transmitted information is identified; based on the transmitted information, it is determined whether the optical communication device is a target optical communication device; and if the optical communication device is a target optical communication device, controlling the A machine that can move autonomously or a part thereof travels to the optical communication device. The method according to claim 7, further comprising: if the optical communication device is not a target optical communication device, identifying the optical communication device based on the information transmitted by the optical communication device, and obtaining the optical communication device and The relative positional relationship between the target optical communication device; determine the relative positional relationship between the autonomously movable machine or its part and the optical communication device; determine the target optical communication device and the autonomously movable The relative positional relationship between the machine or its parts; and based at least in part on the relative positional relationship between the target optical communication device and the autonomously movable machine or its part. Partially guide the target optical communication device. The method according to claim 8, wherein determining the relative positional relationship between the autonomously movable machine or part thereof and the optical communication device includes: The relative positional relationship between the machine or its part that can move autonomously and the optical communication device is determined by relative positioning. The method according to claim 7, wherein collecting information transmitted by a certain optical communication device around and identifying the transmitted information through the rolling shutter camera installed on the machine capable of moving autonomously includes: The shutter camera obtains continuous multiple frames of images of the optical communication device; for each frame of image, determine whether there are fringes in the part of the image corresponding to the position of the light source or what type of fringes exist; and Determine the information represented by each frame of image. The method according to claim 7, further comprising: first controlling the machine capable of autonomous movement to travel near the target optical communication device. The method according to claim 11, wherein first controlling the autonomously movable machine to travel near the target optical communication device comprises: guiding the autonomously movable machine to the target optical communication device at least in part through a satellite navigation system. Near the communication device; and/or at least partially utilize the relative positional relationship between other optical communication devices and the target optical communication device to guide the autonomously movable machine to the vicinity of the target optical communication device. The method according to claim 12, wherein, at least partially using the relative position relationship between the other optical communication device and the target optical communication device to guide the autonomously movable machine to the vicinity of the target optical communication device comprises : The machine capable of autonomous movement recognizes the other optical communication device while traveling, and obtains the relative position relationship between the other optical communication device and the target optical communication device; Determine the relative positional relationship between the autonomously movable machine and the other optical communication device; determine the relative positional relationship between the target optical communication device and the autonomously movable machine; and based at least in part on the The relative positional relationship between the target optical communication device and the autonomously movable machine guides the autonomously movable machine to the vicinity of the target optical communication device. The method according to claim 7, wherein determining whether the optical communication device is a target optical communication device based on the transmitted information includes: determining whether the transmitted information explicitly or implicitly includes predetermined information. The method according to claim 14, wherein the predetermined information is a predetermined identification or verification code. The method according to claim 7, wherein determining whether the optical communication device is a target optical communication device based on the transmitted information comprises: determining whether the optical communication device is a target optical communication device by the machine capable of autonomous movement Or the machine that can move autonomously transmits the transmitted information to the server, and the server determines whether the optical communication device is a target optical communication device based on the transmitted information, and sends the determination result to the A machine that can move autonomously. A machine that can move autonomously, including a rolling shutter camera, a processor, and a memory. The memory stores a computer program. When the computer program is executed by the processor, the computer program can be used to implement items such as request 7-16. The method of any one of. A storage medium in which a computer program is stored, and when the computer program is executed, it can be used to implement the method described in any one of Claims 7-16.

Images