CN114999308B - Map generation method, device, computer readable storage medium and computer equipment - Google Patents

Abstract

The application relates to a map generation method, a map generation device, a computer readable storage medium and a computer device. The method comprises the following steps: starting from any first type of position point of the indoor space, scanning the indoor space; when passing through a first type of position point in the indoor space, acquiring an absolute coordinate corresponding to the first type of position point; acquiring sensor data obtained by scanning the first type of position points and the second type of position points in the indoor space in real time; determining first estimated coordinates of the first type of position points and second estimated coordinates of the second type of position points according to the sensor data; judging whether the indoor space has a dead zone which is not scanned or not; if so, returning to the step of scanning the indoor space starting from any first type of position point of the indoor space; if not, calculating the deviation between the first estimated coordinates and the absolute coordinates; and when the deviation is smaller than a preset threshold value, generating a map according to each second estimated coordinate.

Description

Map generation method, device, computer readable storage medium and computer equipment

Technical Field

The present application relates to the field of computer technology, and in particular to a map generation method, device, computer-readable storage medium and computer equipment.

Background Art

In automated warehouse management, handling robots are becoming more and more popular. As an important tool for automated transportation, handling robots are gradually replacing traditional manual handling and sorting, which can greatly improve the efficiency of cargo transfer and transportation, and automatic loading and unloading of shelves.

Just like manual handling, a handling robot needs to know its own location and the destination when carrying goods. This requires providing the handling robot with a map of the indoor space, such as a warehouse map. Currently, the construction of a map of the indoor space where the handling robot is located is mostly achieved by pasting a large number of QR codes on the ground, which requires a lot of manpower and material resources.

Summary of the invention

Based on this, it is necessary to provide a map generation method, device, computer-readable storage medium and computer equipment to address the existing technical problem of constructing indoor space maps by pasting a large number of QR codes, which consumes a lot of manpower and material resources.

A map generation method, applied to a robot, comprising:

Starting from any first-category position point in the indoor space, scanning the indoor space;

When passing through a first type of position point in the indoor space, obtaining the absolute coordinates corresponding to the first type of position point;

Acquire sensor data obtained by scanning the first type of location points and the second type of location points in the indoor space in real time;

Determine, based on the sensor data, first estimated coordinates of each of the first-category position points and second estimated coordinates of each of the second-category position points;

calculating a deviation between the first estimated coordinate and the absolute coordinate;

When the deviation is less than a preset threshold, a map is generated according to each of the second estimated coordinates.

In one embodiment, the real-time acquisition of sensor data obtained by scanning the first type of location points and the second type of location points in the indoor space includes:

Acquire image data collected by a visual sensor in the lane in real time;

Determining whether there is a second graphic code for locating the second type of position point in the image according to the image data;

If so, obtaining the first offset data of the current robot relative to the first type of position point;

Determine second offset data of the second graphic code relative to the current robot according to the image;

Determining the first estimated coordinates of each of the first-category position points and the second estimated coordinates of each of the second-category position points according to the sensor data comprises:

The estimated coordinates of the second graphic code are calculated according to the first offset data, the absolute coordinates and the second offset data, and the estimated coordinates of the second graphic code are used as the second estimated coordinates.

In one embodiment, judging whether there is a second graphic code for locating the second type of position point in the image according to the image data comprises:

Obtaining image attributes corresponding to the second graphic code;

Extracting image features of the image based on currently acquired image data;

When the image feature matches the image attribute, it is determined that the second graphic code exists in the image.

In one embodiment, the method further comprises:

When there is a blind area in the indoor space that has not been scanned;

Return to the step of scanning the indoor space from any first-category position point in the indoor space; when there is a second-category position point in the lane that has not been scanned,

Return to the step of scanning the indoor space from any first-category position point in the indoor space.

In one embodiment, the sensor data includes data collected by multiple sensors; the method further includes:

When the deviation is greater than a preset threshold, adjusting the trust level corresponding to each sensor;

Return to the step of scanning the indoor space from any first-category position point in the indoor space;

Processing the data collected by the corresponding sensor according to the adjusted confidence level;

Based on the processed data, first estimated coordinates of each of the first-category position points and second estimated coordinates of each of the second-category position points are determined.

In one embodiment, after generating a map according to each of the first estimated coordinates when the deviation is less than a preset threshold, the method further includes:

Receiving target coordinates of a target location to be traveled to;

During the journey, obtain real-time data from scanning the current location;

Determine the coordinates of the current location according to the real-time data;

Go to the target location according to the coordinates of the current location and the target coordinates.

In one embodiment, when passing through a first type of location point in the indoor space, obtaining the absolute coordinates corresponding to the first type of location point includes:

Acquire images corresponding to the first type of location points;

The first graphic code in the image is parsed to obtain the absolute coordinates of the tunnel entrance.

In one embodiment, the first type of location points includes known location points at the entrances of lanes in the indoor space; the second type of location points includes location points corresponding to each storage location on the shelves in the lanes in the indoor space and/or location points corresponding to the public areas.

A map generating device, comprising:

A scanning module, used to scan the indoor space from any first type position point in the indoor space;

An absolute coordinate acquisition module, used for acquiring the absolute coordinates corresponding to the first type of position points in the indoor space when passing through the first type of position points;

A sensor data acquisition module, used for acquiring in real time sensor data obtained by scanning the first type of location points and the second type of location points in the indoor space;

an estimated coordinate determination module, configured to determine, based on the sensor data, a first estimated coordinate of each of the first-category position points and a second estimated coordinate of each of the second-category position points;

a deviation calculation module, used for calculating the deviation between the first estimated coordinate and the absolute coordinate;

A map generation module is used to generate a map according to each of the second estimated coordinates when the deviation is less than a preset threshold.

A computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the processor executes the steps of the above-mentioned map generation method.

A computer device comprises a memory and a processor, wherein the memory stores a computer program, and when the computer program is executed by the processor, the processor executes the steps of the above-mentioned map generation method.

The above-mentioned map generation method, device, computer-readable storage medium and computer equipment, when generating a map, start from any first-class position point in the indoor space and scan the indoor space; when passing through the first-class position point in the indoor space, obtain the absolute coordinates corresponding to the first-class position point; and calculate the deviation between the second estimated coordinates and the absolute coordinates; when the deviation is less than the preset threshold, it means that the sensor data collected by the sensor is sufficiently reliable, so the estimated coordinates of each position point determined based on these sensors are also reliable, so that a map can be generated based on the estimated coordinates of these position points, thereby ensuring the accuracy of the generated map. Compared with generating a map by pasting a large number of QR codes, which requires a lot of manpower and material resources, this solution can also greatly save manpower and material resources.

BRIEF DESCRIPTION OF THE DRAWINGS

FIG1 is an application environment diagram of a map generation method in one embodiment;

FIG2 is a schematic diagram of a flow chart of a map generation method in one embodiment;

3 is a schematic diagram of a process of determining whether a second graphic code for locating a second type of position point exists in an image according to image data in one embodiment;

FIG4 is a schematic diagram of a process of generating a map in one embodiment;

FIG5 is a schematic diagram of a process of navigating according to a generated map in one embodiment;

FIG6 is a schematic diagram of a process of generating a map when a known location point is represented by a QR code in one embodiment;

FIG7 is a schematic diagram of a flow chart of a map generation method in another embodiment;

FIG8 is a structural block diagram of a map generating device in one embodiment;

FIG9 is a structural block diagram of a map generating device in another embodiment;

FIG10 is a structural block diagram of a map generating device in yet another embodiment;

FIG. 11 is a block diagram of a computer device in one embodiment.

DETAILED DESCRIPTION

In order to make the purpose, technical solution and advantages of the present application more clearly understood, the present application is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application.

It is understood that the terms "first", "second", etc. used in the present invention may be used herein to describe various elements, but these elements are not limited by these terms. These terms are only used to distinguish a first element from another element. For example, a first type of location point may be referred to as a second type of location point, and similarly, a second type of location point may be referred to as a first type of location point without departing from the scope of the present invention. Both the first type of location point and the second type of location point are type of location points, but they are not the same type of location points.

FIG. 1 is an application environment diagram of a map generation method in an embodiment. Referring to FIG. 1 , the map generation method is applied to a map generation system. The map generation system includes a robot 110 and a server 120. The terminal 110 and the server 120 are connected via a network. The robot 110 may also be an automatic guided vehicle, which may be used to transport goods according to the generated map. The server 120 may be implemented as an independent server or a server cluster consisting of multiple servers.

Taking the first type of location point as a known location point at the entrance of the lane in the indoor space, and the second type of location point as a location point corresponding to each storage location on the shelf in the lane in the indoor space as an example, the robot 110 can start from any known location point in the indoor space, scan the indoor space, and obtain the sensor data obtained by scanning each storage location on the shelf in the lane in the indoor space in real time, determine the estimated coordinates of each storage location based on the sensor data, and when passing through the lane entrance, obtain the absolute coordinates corresponding to the lane entrance, calculate the deviation between the estimated coordinates and the absolute coordinates at the lane entrance, and when the deviation is less than the preset threshold, generate a map based on each estimated coordinate. The robot 110 can also report the obtained sensor data to the server 120, and the server 120 determines the estimated coordinates of each storage location on the shelf in the lane according to the sensor data, calculates the deviation between the absolute coordinates at the lane entrance and the estimated coordinates, and when the calculated deviation is less than the preset threshold, the server 120 can generate a map based on the estimated coordinates of each storage location.

As shown in FIG2 , in one embodiment, a map generation method is provided. This embodiment mainly uses the method applied to the robot 110 in FIG1 as an example. Referring to FIG2 , the map generation method specifically includes the following steps:

S202, starting from any first-category position point in the indoor space, scanning the indoor space.

Among them, the indoor space is the space for which the map is to be generated, such as an indoor warehouse. The process of the robot scanning the indoor space is the process of the robot collecting data on the indoor space through various sensors during the movement.

The indoor space includes a first type of position point, which is a feature point with given absolute coordinates, which can be called a known position point. The known position point can be the position point at the entrance of the lane mentioned above, that is, the position points at the entrance of the lane can all be known position points, and the absolute coordinates are the coordinates pre-set for the known position point. In the indoor space, in order to save the cost of manually measuring feature points, the number of known position points can be a small number, such as two or three. The robot can start from any known position point and scan the indoor space through various sensors. In order to ensure the accuracy of the map in the lane, a known position point can be set at the entrance of the lane. For example, in the indoor space, the known position points may include the entrance of Lane 1, the entrance of Lane 2, or the starting point C of the main road, and so on. In some embodiments, a QR code or bar code can be pasted at the known position point, and the absolute coordinates of the known position point can be obtained by parsing the scanned QR code or bar code.

S204, when passing through a first type of position point in the indoor space, obtaining the absolute coordinates corresponding to the first type of position point.

In one embodiment, the first type of position point includes a position point at a lane entrance in an indoor space and/or a known position point. Taking the first type of position point as a position point at a lane entrance in an indoor space as an example, when passing through the lane entrance of the first type of position point, the absolute coordinates corresponding to the lane entrance of the first type of position point are obtained.

Specifically, when the robot passes through the lane entrance according to the collected image data during the movement, the pre-set absolute coordinates of the lane entrance can be obtained. For example, the shelves at each lane entrance in the indoor space can be used as known position points, and the corresponding absolute coordinates can be given to the shelves at the lane entrance. When the robot passes through these lane entrances during the movement, the corresponding absolute coordinates of the lane entrance can be obtained, and the estimated coordinates of each storage location can be corrected with the absolute coordinates to obtain a map with higher accuracy.

S206, acquiring sensor data obtained by scanning the first type of location points and the second type of location points in the indoor space in real time.

In one embodiment, the second type of location points include location points corresponding to each storage location on the shelves in the lanes in the indoor space and/or location points corresponding to the public area. Taking the second type of location points as location points corresponding to each storage location on the shelves in the lanes as an example, the robot can obtain sensor data obtained by scanning each storage location on the shelves in the lanes in the indoor space in real time.

Among them, sensor data is data collected by various sensors configured for the robot. The configured sensors can measure the deviation direction and offset distance of the robot's current state relative to the initial state, and use the deviation direction and offset distance as the sensor data of the current robot scanning the indoor space. Sensors may include odometers, and IMUs (inertial measurement units) composed of gyroscopes, accelerometers, compasses, etc., and may also include laser rangefinders, visual sensors for collecting image data, etc.

The sensor data may also include image data collected by a visual sensor, which may be a camera, and the robot may perform image processing on the collected image data to determine the various stock locations on the shelves in the aisle. The image data and the offset data measured by the IMU may be combined to determine the stock locations in the map to be generated and the coordinates corresponding to the stock locations.

In one embodiment, the robot can collect image data through a first sensor and collect offset data through a second sensor. The first sensor can be a visual sensor, such as a camera, etc., and the second sensor can be various measurement sensors constituting an IMU. It should be noted that the "first" and "second" here are not used to limit the number of sensors, but are only used to distinguish sensors that collect different data.

The storage location is the pickup point on the shelf in the lane, for example, storage location 1A01 on shelf 1A in lane 1, storage location 1B02 on shelf 1B in lane 1, and storage location 2A01 on shelf 2A in lane 2, etc. The process of generating the map in the lane is the process of determining the coordinates of these storage locations. After having a map including these storage locations, when the robot moves in the indoor space, it can determine the direction and distance to be moved next according to the coordinates of these storage locations, thereby realizing map navigation. The coordinates of the storage location can be two-dimensional plane coordinates or three-dimensional spatial coordinates. In the process of scanning the indoor space, the robot obtains the data collected by all current sensors in real time. When scanning the storage location on the shelf, it also obtains the sensor data of the current scanning storage location.

S208: Determine, according to the sensor data, first estimated coordinates of each first-category position point and second estimated coordinates of each second-category position point.

The estimated coordinates are calculated based on the sensor data collected by the robot during its movement. Since the data collected by the sensor is noisy and may not be very accurate, or the trust in the sensor data is not high enough when calculating based on the sensor data, the calculated coordinates can be called estimated coordinates.

Specifically, taking the example of determining the estimated coordinates of each storage location based on sensor data, the robot can initialize the status of each sensor, and after initializing each sensor, scan the indoor space from a known position point, and obtain sensor data during the scanning process. When the robot determines that the storage location has been scanned based on the image data, the estimated coordinates of the storage location relative to the known position point are determined based on the current offset data. The known position point can be set at the entrance of the lane.

In one embodiment, when the robot detects the next known location point during the scanning process, the robot can initialize the status of each sensor again. After initialization, it continues to scan the indoor space. When a storage location on the shelf is detected, the estimated coordinates of the storage location can be calculated based on this known location point and the current sensor data.

For example, when the robot starts from any known position, the initial coordinates of the X-axis and Y-axis are S(0,0), and the initial angle measured by the robot at the initial coordinates is 0 degrees. During the movement, when the robot scans a certain storage location, the corresponding offset data is 5 meters and 90 degrees, then the estimated coordinates of the storage location are calculated to be (0,5). In some embodiments, the estimated coordinates can be two-dimensional plane coordinates or three-dimensional spatial coordinates.

S210: Calculate the deviation between the first estimated coordinates and the absolute coordinates.

Taking the entrance of the tunnel as an example, the robot can calculate the deviation between the estimated coordinates at the entrance of the tunnel and its absolute coordinates. Specifically, when the robot passes through the entrance of the tunnel, it obtains the corresponding absolute coordinates and calculates the deviation between the estimated coordinates at the entrance of the tunnel and the absolute coordinates. In one embodiment, the deviation between the estimated coordinates and the absolute coordinates can be calculated by the Euclidean distance formula. For example, when the estimated coordinates and the absolute coordinates are both two-dimensional coordinates, the calculated estimated coordinates are (0.96, 1.05), and the given absolute coordinates are (1, 1), and the units are all millimeters. Then the deviation is (mm).

S212: When the deviation is less than a preset threshold, a map is generated according to each second estimated coordinate.

Specifically, the absolute coordinates at the entrance of the alley can be used to make closed-loop corrections to the estimated coordinates of all storage locations. When the deviation between the estimated coordinates and the absolute coordinates at the entrance of the alley is small, it means that the estimated coordinates calculated based on the sensor data are closer to the absolute coordinates, the noise in the sensor data is lower, and the accuracy of the sensor collected data is higher. Then, the accuracy of the estimated coordinates of each storage location calculated based on the sensor data is also higher.

In one embodiment, the deviation less than the preset threshold may mean that the deviation corresponding to each lane entrance is less than the set threshold. That is, when the robot determines that the deviation corresponding to each lane entrance is less than the threshold, it can output a map based on the estimated coordinates of each storage location calculated during the scanning process. The deviation less than the preset threshold may also mean that the robot calculates the total deviation based on the deviations of all lane entrances and is less than the set threshold. For example, there are n lane entrances with known absolute coordinates in an indoor warehouse, and the deviation corresponding to each lane entrance is E _i (i=1,2,3,...,n). Then the total deviation can be expressed as the average of the sum of the deviations: Only when the total deviation is less than the preset threshold, the robot can output a map based on the calculated estimated coordinates of each storage location.

The above-mentioned map generation method, when generating a map, starts from any first-class position point in the indoor space and scans the indoor space; when passing through the first-class position point in the indoor space, the absolute coordinates corresponding to the first-class position point are obtained; and the deviation between the second estimated coordinates and the absolute coordinates is calculated; when the deviation is less than the preset threshold, it means that the sensor data collected by the sensor is sufficiently reliable, so the estimated coordinates of each position point determined based on these sensors are also reliable, and thus a map can be generated based on the estimated coordinates of these position points, thereby ensuring the accuracy of the generated map. In addition, compared with generating a map by pasting a large number of QR codes, which requires a lot of manpower and material resources, the above-mentioned map generation method can also greatly save manpower and material resources.

In one embodiment, step S204 includes: collecting an image corresponding to the first type of location point; parsing the first graphic code in the image to obtain the absolute coordinates of the tunnel entrance.

Taking the first type of position point as the position point at the entrance of the lane as an example, the robot can collect images on the shelves at the entrance of the lane; and parse the first graphic code in the image to obtain the absolute coordinates of the entrance of the lane.

The first graphic code is a marking image containing positioning information. The first graphic code may be a QR code or a barcode. The known position points set in the indoor space may be marked with the first graphic code, so that the absolute coordinates corresponding to the entrance of the lane can be obtained by parsing the first graphic code. For example, the first graphic code may be pasted on the side of the shelf at the entrance of the lane. When the robot passes through the entrance of the lane, the absolute coordinates corresponding to the entrance of the lane may be obtained by collecting the first graphic code and parsing the image code.

In one embodiment, step S206 includes: acquiring image data collected by the visual sensor in the lane in real time; judging whether there is a second graphic code for locating the second type of position point in the image according to the image data; if there is, acquiring the first offset data of the current robot relative to the first type of position point; determining the second offset data of the second graphic code relative to the current robot according to the image. Step S208 includes: calculating the estimated coordinates of the second graphic code according to the first offset data, the absolute coordinates and the second offset data, and using the estimated coordinates of the second graphic code as the second estimated coordinates.

Similarly, using the storage location and the alley entrance as an example, the robot can obtain image data collected by the visual sensor in the alley in real time; determine whether there is a second graphic code in the image based on the image data; if so, obtain the first offset data of the current robot relative to the alley entrance; determine the second offset data of the second graphic code relative to the current robot based on the image; calculate the estimated coordinates of the second graphic code based on the first offset data, the absolute coordinates corresponding to the alley entrance and the second offset data, as the estimated coordinates of the storage location.

Among them, the image data includes the image of the indoor space collected by the robot during the scanning process. The offset data includes the deviation direction, the offset distance, etc. The first offset data is the offset data of the current robot relative to the entrance of the lane. The first offset data can be the data collected by the sensor when the robot travels from the entrance of the lane to the current position, or it can be the intermediate data calculated based on the data collected when the robot travels from a known position point to the entrance of the lane and the data collected when it travels from a known position point to the current position. The second offset data is the offset data of the second graphic code relative to the visual sensor of the current robot. The second offset data can be calculated based on the pixel width size and the actual width size of the second graphic code in the collected image. The second graphic code is a marking code pasted on each storage location, which is used to locate each storage location, and can be a storage location QR code or a storage location barcode. In some embodiments, other iconic items can also be used to mark each storage location on the shelf in the lane.

Specifically, when the robot moves in the lane, it can collect image data through the camera, and the robot can determine whether the second graphic code exists based on the image data. If it exists, the robot can calculate the distance between the second graphic code and the camera of the current robot based on the actual width of the second graphic code, the focal length of the camera, and the pixel width of the second graphic code in the image, and then calculate the height of the second graphic code relative to the camera of the current robot based on the distance and the torsion angle of the camera. According to the distance and height, the second offset data of the second graphic code relative to the current robot at various angles can be obtained. Further, the robot determines the estimated coordinates of the current robot based on the currently collected first offset data and the absolute coordinates corresponding to the lane entrance, and then calculates the estimated coordinates of the second graphic code based on the second offset data and the estimated coordinates of the current robot as the estimated coordinates of the storage location corresponding to the second graphic code.

It should be noted that if the robot's current estimated coordinates are two-dimensional plane coordinates, the second offset data may be the offset information of the second graphic code relative to the current robot on the plane; if the estimated coordinates are three-dimensional spatial coordinates, the second offset data may be the offset information of the second graphic code relative to the current robot in space.

For example: when the robot moves from the entrance of the lane to the second graphic code, the corresponding second offset data is 2.5 meters and 90 degrees. If the absolute coordinates corresponding to the entrance of the lane are (0,0), the estimated coordinates of the current robot are (0,2.5). The robot calculates the second offset data of the second graphic code relative to the current robot based on the pixel size of the second graphic code in the image, the torsion angle of the camera, and the actual size of the second graphic code. For example, if it is determined that the second graphic code is 0.3 meters in the negative direction of the current robot's X-axis, the estimated coordinates of the second graphic code on the plane are calculated to be (-0.3,2.5). If the height of the robot camera is 1.6 meters, the second graphic code is determined to be 0.4 meters below the robot camera based on the torsion angle, and the estimated coordinates of the second graphic code in space are calculated to be (-0.3,2.5,1.2).

Similarly, the robot can determine the first estimated coordinates of the first type of position point currently passed by in the above manner: the robot obtains image data collected by the visual sensor in real time; determines whether there is a first graphic code in the image based on the image data; if so, obtains the offset data of the current robot relative to the first type of position point passed last time; determines the offset data of the first graphic code relative to the current robot based on the image; calculates the estimated coordinates of the first graphic code based on the absolute coordinates corresponding to the first type of position point passed last time and the two offset data, as the estimated coordinates of the first type of position point currently passed.

By calculating the corresponding first estimated coordinates and second estimated coordinates using the first offset data, the second offset data and the absolute coordinates corresponding to the first type of position points, the accuracy of the estimated coordinate calculation can be improved.

As shown in FIG3 , in one embodiment, judging whether there is a second graphic code for locating a second type of position point in the image according to the image data includes:

S302: Obtain image attributes corresponding to the second graphic code.

Among them, the second graphic code corresponds to the storage location on the shelf. When locating the storage location on the shelf, the second graphic code can be used to determine whether the storage location has been scanned. When scanning in the alley, the camera will scan the shelves, the ground, the second graphic code, etc. in the alley. The robot needs to determine whether the second graphic code exists from the collected image data. The image attributes include the color characteristics and brightness characteristics of each pixel in the image, and can also include the depth characteristics of the pixel points. The image attributes of the second graphic code can be analyzed in advance, and the second graphic code can be stored in correspondence with the corresponding image attributes. In this way, the robot can obtain the graphic attributes corresponding to the second graphic code and can determine that the collected image includes the second graphic code.

S304: extracting image features of the image based on the currently collected image data.

S306: When the image feature matches the image attribute, it is determined that a second graphic code exists in the image.

Specifically, after the robot acquires the image data, it can extract the image features of the image from the acquired image data in the same manner as extracting the image attributes, and match the extracted image features with the pre-stored image attributes to determine whether the currently acquired image data includes the second graphic code. When the extracted image features are successfully matched with the pre-stored image attributes, it can be determined that the second graphic code exists in the image acquired by the robot at the current position.

In one embodiment, the map generation method further includes: acquiring in real time third offset data obtained by scanning the common area in the indoor space; and determining the estimated coordinates of the robot's current position in the common area based on the third offset data and known position points.

The indoor space includes a common area and an aisle area for inventory, and the aisle area is formed by the space area between adjacent shelves. Specifically, the estimated coordinates of the current robot position can also be calculated based on the third offset data of the current robot relative to the known position point when scanning the common area, and the absolute coordinates corresponding to the known position point. The third offset data obtained by the robot when scanning the common area can be collected by IMU.

In one embodiment, the map generation method also includes: when passing through a known location point, obtaining the absolute coordinates corresponding to the known location point; calculating the second deviation between the estimated coordinates of the known location point and the corresponding absolute coordinates; when the deviation is less than a preset threshold, generating a map based on each estimated coordinate includes: when the second deviation is less than a preset threshold, generating a map based on each estimated coordinate.

In this embodiment, when the robot is scanning in the common area, each estimated coordinate can be corrected in a closed loop through the known position points in the common area, thereby improving the accuracy of the common area map. Specifically, when the robot passes through a known position point during scanning in the common area, the estimated coordinates of the known position point can be calculated based on the estimated coordinates of the previous position point and the offset data of the current robot relative to the previous position point. During the correction, the second deviation between the estimated coordinates of the known position point and the corresponding absolute coordinates can be calculated. When the second deviation is less than the preset threshold, it means that the data error collected by each sensor on the robot is small, and a map can be generated based on the estimated coordinates of each storage location on the shelf in the aisle.

In one embodiment, the method also includes: when there is a blind spot in the indoor space that has not been scanned; returning to the step of scanning the indoor space from any first-class position point in the indoor space; when there is a storage location in the alley that has not been scanned, returning to the step of scanning the indoor space from any first-class position point in the indoor space.

Among them, blind spots refer to areas that have not been scanned when the robot scans the indoor space. Specifically, if there are blind spots that have not been scanned in the indoor space, it is necessary to continue scanning to ensure that the generated map is complete enough. If there are storage locations that have not been scanned in the indoor space, it is also necessary to continue scanning to ensure that the coordinates of each storage location in the obtained map are accurate enough.

In one embodiment, the robot can count the categories to which all position points detected during the scanning of the indoor space belong, obtain the preset categories of all position points set for the indoor space, and determine whether the categories to which all detected position points belong completely cover the preset categories. If not, it means that there are still blind spots that have not been scanned in the current indoor space, and the robot needs to return to execute the steps of scanning the indoor space from any known position point in the indoor space, and continue scanning the indoor space until the categories of the scanned position points completely cover the preset categories.

Furthermore, when the categories of the location points scanned by the robot in the indoor space completely cover the preset categories, the number of scanned storage locations is counted to obtain the total number of storage locations in the indoor space, and the counted number is compared with the total number to determine whether there are storage locations in the indoor space that have not been scanned. When there are still storage locations in the indoor space that have not been scanned, the robot needs to return to execute the step of scanning the indoor space from any known location point in the indoor space, and continue to scan the indoor space until all storage locations are scanned.

As shown in FIG4 , it is a schematic diagram of a process of generating a map in an embodiment. Referring to FIG4 , firstly, several known position points with given absolute coordinates are set in the indoor space where the map is to be generated, the robot starts from any known position point and scans the indoor space, obtains sensor data during the scanning process, obtains the estimated coordinates of each storage location according to the sensor data, and when the robot passes through the entrance of the lane, calculates and records the deviation between the estimated coordinates and the absolute coordinates at the entrance of the lane, and the robot determines whether there is a blind spot that has not been scanned in the current indoor space, if so, returns to the step of scanning the indoor space from any known position point in the indoor space, if not, the robot determines whether there is a storage location that has not been scanned in the current environment indoor space, if so, returns to the step of scanning the indoor space from any known position point in the indoor space, if not, the robot determines whether the deviation corresponding to each recorded entrance of the lane is less than a threshold, if not, returns to the step of scanning the indoor space from any known position point in the indoor space, if so, outputs a map according to the estimated coordinates of each storage location finally obtained.

In one embodiment, the sensor data includes data collected by multiple sensors; the map generation method also includes: when the deviation is greater than a preset threshold, adjusting the trust level corresponding to each sensor; returning to the step of scanning the indoor space from any first-class position point in the indoor space; processing the data collected by the corresponding sensor according to the adjusted trust level; and determining the estimated coordinates of each storage location based on the processed data.

The sensor data obtained by the robot when scanning the indoor space can be obtained by fusing the data collected by each sensor. The accuracy of the estimated coordinates of each storage location calculated can be improved by processing the sensor data obtained by data fusion. Due to the influence of factors such as sensor accuracy and noise data, the accuracy of the sensor data collected by the robot through each sensor when scanning the indoor space will also be different. The trust degree is the degree of trust between other sensor data and the data collected by the current sensor. In one embodiment, the trust degree can be represented by the weight of the data collected by the current sensor in the fusion process. The larger the weight value, the higher the authenticity of the data collected by the current sensor.

Specifically, when the deviation between the estimated coordinates calculated based on the sensor data and the actual coordinates is greater than a preset threshold, the trust in the parameters of each sensor can be adjusted. After the adjustment, the step of scanning the indoor space from any known position point in the indoor space is returned to the execution again, and the data collected during this scanning process is fused according to the adjusted trust to obtain new sensor data, and the estimated coordinates of each position point and storage location are determined based on the new sensor data.

Accordingly, the robot may calculate again the deviation between the estimated coordinates and the absolute coordinates at the entrance of the lane to see whether the deviation is less than a threshold value, so as to determine whether a map can be generated based on the estimated coordinates obtained this time.

In this embodiment, when the deviation is greater than the threshold, the trustworthiness of the data collected by each sensor may be adjusted to improve the accuracy of the estimated coordinates of each feature point.

In one embodiment, as shown in FIG5 , the map generation method further includes the step of navigating according to the generated map, specifically including:

S502, receiving the target coordinates of the target position to be traveled to.

The target location can be a target warehouse. After the customer successfully places an order, the order system can determine the identifier used to mark the goods according to the goods information in the order. The identifier can be, for example, a single product identifier (SKU). According to the identifier, the warehouse location of the goods in the warehouse is determined as the target warehouse location, and the coordinates corresponding to the target warehouse location are sent to the robot. The robot can receive the target coordinates sent by the order system.

S504, during the moving process, obtaining real-time data obtained by scanning the current position.

After receiving the target coordinates, the robot starts from the starting position and scans the indoor space while moving to obtain real-time data collected by the sensor. The real-time data includes image data and offset data obtained by fusing the data collected by the current sensor.

S506, determining the coordinates of the current location according to the real-time data.

During the movement, the robot can calculate the coordinates of the current position in real time based on the collected real-time data and the coordinates of the starting point, and compare the coordinates of the current position with the target coordinates.

S508, proceed to the target location according to the coordinates of the current location and the target coordinates.

After the robot obtains the coordinates of its current location, it knows where it is in the map, compares the coordinates of its current location with the target coordinates, and thus determines its next route, and follows the route to the target coordinates and arrives at the target storage location.

In this embodiment, after a map corresponding to the current environment is generated, the robot can navigate according to the map and the currently scanned sensor data to move to the target storage location.

As shown in FIG6 , it is a schematic diagram of a process of generating a map when a known position point is represented by a QR code in an embodiment. Referring to FIG6 , firstly, a QR code is pasted at a known position point (including an aisle entrance) according to a given absolute coordinate, and these QR codes can be parsed to obtain absolute coordinates; second graphic codes are pasted on the side of each storage location on the shelf, and these second graphic codes are only used to represent each storage location; then the robot moves in the indoor space, and during the movement, the estimated coordinates of each position point are calculated according to the currently collected sensor data; when the robot reaches the aisle entrance, the QR code at the aisle entrance is parsed to obtain absolute coordinates, and the deviation is calculated according to the absolute coordinates obtained by parsing and the estimated coordinates at the aisle entrance; the sensor data is calibrated according to the deviation, and the estimated coordinates of the storage location and other position points are recalculated, until all storage locations are scanned and the deviation between the estimated coordinates at the aisle entrance and the absolute coordinates is less than a preset threshold, then the map is output according to the estimated coordinates of the second graphic codes pasted on each storage location.

In one embodiment, as shown in FIG7 , the map generation method specifically includes the following steps:

S702, starting from any known position point in the indoor space, scanning the indoor space.

S704, when passing through the tunnel entrance, obtaining the absolute coordinates corresponding to the tunnel entrance.

S706, acquiring image data collected by the visual sensor in the lane in real time.

S708: Obtain image attributes corresponding to the second graphic code.

S710, extracting image features of an image based on currently collected image data; when the image features match image attributes, determining that a second graphic code exists in the image.

S712, obtaining first offset data of the current robot relative to the lane entrance; and determining second offset data of the second graphic code relative to the current robot according to the image.

S714, calculating the estimated coordinates of the second graphic code according to the first offset data, the absolute coordinates corresponding to the lane entrance and the second offset data as the estimated coordinates of the storage location.

S716, calculating the deviation between the estimated coordinates and the absolute coordinates at the tunnel entrance.

S718, acquiring in real time third offset data obtained by scanning the public area in the indoor space; and determining the estimated coordinates of the current position of the robot in the public area according to the third offset data and the known position points.

S720: When passing through a known position point, calculate a second deviation between the estimated coordinates of the known position point and the corresponding absolute coordinates.

S722, when the deviation between the estimated coordinates and the absolute coordinates at the tunnel entrance or the second deviation between the estimated coordinates and the corresponding absolute coordinates of the known position point is less than a preset threshold, generate a map according to the estimated coordinates.

S724, determine whether there is a blind spot in the indoor space that has not been scanned; if so, return to step 702; if not, execute step S726.

S726, determine whether there are any storage locations in the lane that have not been scanned, if so, return to step 702, if not, execute step S728.

S728, when determining whether the deviation or the second deviation is less than a preset threshold, if so, execute step S732; if not, execute step S730.

S730, adjusting the trust level corresponding to each sensor, and processing the data collected by the corresponding sensor according to the adjusted trust level; and determining the estimated coordinates of each storage location based on the processed data.

S732, receiving the target coordinates of the target storage location to be moved to.

S734, during the moving process, obtain real-time data obtained by scanning the current position; and determine the coordinates of the current position according to the real-time data.

S736, proceed to the target storage location according to the coordinates of the current location and the target coordinates.

The above-mentioned map generation method, when generating a map, starts from any known position point in the indoor space, scans the indoor space, and uses the sensor data obtained by scanning each storage location on the shelf in the aisle of the indoor space to automatically calculate the estimated coordinates of each storage location on the shelf. When the estimated coordinates calculated at the entrance of the aisle are close enough to the given absolute coordinates, it means that the sensor data collected by the sensor is reliable enough. In this way, the estimated coordinates of each storage location on the shelf determined based on these sensors are also reliable. A map can be generated based on the estimated coordinates of these storage locations. Compared with generating a map by pasting a large number of QR codes, which requires a lot of manpower and material resources, it can greatly save manpower and material resources.

It should be understood that, although each step in the above-mentioned each flow chart is shown in turn according to the indication of the arrow, these steps are not necessarily performed in turn according to the order indicated by the arrow. Unless there is clear explanation in this article, the execution of these steps does not have strict order restriction, and these steps can be performed in other orders. Moreover, at least a part of the steps in the above-mentioned each embodiment can include a plurality of sub-steps or a plurality of stages, and these sub-steps or stages are not necessarily performed at the same time, but can be performed at different times, and the execution order of these sub-steps or stages is not necessarily performed in turn, but can be performed in turn or alternately with at least a part of other steps or sub-steps or stages of other steps.

As shown in FIG8 , a map generation device 800 is provided, which includes a scanning module 802, an absolute coordinate acquisition module 804, a sensor data acquisition module 806, an estimated coordinate determination module 808, a deviation calculation module 810 and a map generation module 812. Among them:

The scanning module 802 is used to scan the indoor space starting from any first-class position point in the indoor space; the absolute coordinate acquisition module 804 is used to obtain the absolute coordinates corresponding to the first-class position points when passing through the first-class position points in the indoor space; the sensor data acquisition module 806 is used to obtain the sensor data obtained by scanning the first-class position points and the second-class position points in the indoor space in real time; the estimated coordinate determination module 808 is used to determine the first estimated coordinates of each first-class position point and the second estimated coordinates of each second-class position point according to the sensor data; the deviation calculation module 810 is used to calculate the deviation between the first estimated coordinates and the absolute coordinates; the map generation module 812 is used to generate a map based on each second estimated coordinate when the deviation is less than a preset threshold.

In one embodiment, the sensor data acquisition module 806 is also used to acquire image data collected by the visual sensor in the lane in real time; determine whether there is a second graphic code for locating the second type of position point in the image based on the image data; if so, obtain the first offset data of the current robot relative to the first type of position point; and determine the second offset data of the second graphic code relative to the current robot based on the image.

The estimated coordinate determination module 808 is further configured to calculate the estimated coordinates of the second graphic code according to the first offset data, the absolute coordinates and the second offset data, and use the estimated coordinates of the second graphic code as the second estimated coordinates.

In one embodiment, the sensor data acquisition module 806 is further used to acquire image attributes corresponding to the second graphic code; extract image features of the image based on the currently collected image data; and determine that the second graphic code exists in the image when the image features match the image attributes.

In one embodiment, the scanning module 802 is also used for the step of scanning the indoor space from any first-category position point in the indoor space when there is a blind spot in the indoor space that has not been scanned; and when there is a second-category position point in the alley that has not been scanned, returning to the step of scanning the indoor space from any first-category position point in the indoor space.

In one embodiment, as shown in FIG9 , the sensor data includes data collected by multiple sensors; the above-mentioned device further includes:

The sensor data correction module 814 is used to adjust the trust level corresponding to each sensor when the deviation is greater than a preset threshold;

The estimated coordinate determination module 808 is further used to process the data collected by the corresponding sensor according to the adjusted confidence level; and determine the first estimated coordinates of each first-category position point and the second estimated coordinates of each second-category position point based on the processed data.

In one embodiment, as shown in FIG10 , the device further includes:

The navigation module 816 is used to receive the target coordinates of the target position to be traveled to; during the travel process, obtain the real-time data obtained by scanning the current position; determine the coordinates of the current position according to the real-time data; and travel to the target position according to the coordinates of the current position and the target coordinates.

In one embodiment, the absolute coordinate acquisition module 804 is further used to collect an image corresponding to the first type of location point; and parse the first graphic code in the image to obtain the absolute coordinates of the lane entrance.

When generating a map, the above-mentioned map generating device 800 starts from any known position point in the indoor space, scans the indoor space, and automatically calculates the estimated coordinates of each storage location on the shelf by using the sensor data obtained by scanning each storage location on the shelf in the aisle of the indoor space. When the estimated coordinates calculated at the entrance of the aisle are close enough to the given absolute coordinates, it means that the sensor data collected by the sensor is reliable enough. In this way, the estimated coordinates of each storage location on the shelf determined based on these sensors are also reliable, and a map can be generated based on the estimated coordinates of these storage locations. Compared with generating a map by pasting a large number of QR codes, which requires a lot of manpower and material resources, it can greatly save manpower and material resources.

FIG11 shows an internal structure diagram of a computer device in one embodiment. The computer device may specifically be the robot 110 in FIG1 . As shown in FIG11 , the computer device includes a processor, a memory, a network interface, and a sensor connected via a system bus. Among them, the sensor includes a visual sensor, and the memory includes a non-volatile storage medium and an internal memory. The non-volatile storage medium of the computer device stores an operating system and may also store a computer program, which, when executed by the processor, enables the processor to implement a map generation method. The internal memory may also store a computer program, which, when executed by the processor, enables the processor to execute a map generation method.

Those skilled in the art will understand that the structure shown in FIG. 11 is merely a block diagram of a partial structure related to the solution of the present application, and does not constitute a limitation on the computer device to which the solution of the present application is applied. The specific computer device may include more or fewer components than shown in the figure, or combine certain components, or have a different arrangement of components.

In one embodiment, the map generating apparatus 800 provided by the present application may be implemented in the form of a computer program, which may be run on a computer device as shown in FIG11. The memory of the computer device may store various program modules constituting the map generating apparatus 800, and the computer program composed of various program modules enables the processor to execute the steps of the map generating method of various embodiments of the present application described in this specification.

For example, the computer device shown in Figure 11 can execute step S202 through the scanning module 802 in the map generating device 800 shown in Figure 8; execute step S204 through the absolute coordinate acquisition module 804; execute step S206 through the sensor data acquisition module 806; execute step S208 through the estimated coordinate determination module 808; execute step S210 through the deviation calculation module 810; and execute step S212 through the map generating module 812.

In one embodiment, a computer device is provided, including a memory and a processor, wherein the memory stores a computer program, and when the computer program is executed by the processor, the processor executes the steps of the above-mentioned map generation method. The steps of the map generation method here may be the steps of the map generation method in the above-mentioned various embodiments.

In one embodiment, a computer readable storage medium is provided, which stores a computer program, and when the computer program is executed by a processor, the processor executes the steps of the above-mentioned map generation method. The steps of the map generation method here can be the steps of the map generation method in each of the above-mentioned embodiments.

Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiments can be completed by instructing the relevant hardware through a computer program, and the program can be stored in a non-volatile computer-readable storage medium. When the program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to memory, storage, database or other media used in the embodiments provided in the present application can include non-volatile and/or volatile memory. Non-volatile memory may include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM) or flash memory. Volatile memory may include random access memory (RAM) or external cache memory. As an illustration and not limitation, RAM is available in many forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link (Synchlink) DRAM (SLDRAM), memory bus (Rambus) direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM).

The technical features of the above embodiments may be combined arbitrarily. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

The above-mentioned embodiments only express several implementation methods of the present application, and the descriptions thereof are relatively specific and detailed, but they cannot be understood as limiting the scope of the present application. It should be pointed out that, for a person of ordinary skill in the art, several variations and improvements can be made without departing from the concept of the present application, and these all belong to the protection scope of the present application. Therefore, the protection scope of the present application shall be subject to the attached claims.

Claims (10)

1. A map generation method, applied to a robot, comprising: Starting from any first-category position point in the indoor space, scanning the indoor space; When passing through a first type of position point in the indoor space, obtaining the absolute coordinates corresponding to the first type of position point; Acquire sensor data obtained by scanning the first type of location points and the second type of location points in the indoor space in real time; Determine, based on the sensor data, first estimated coordinates of each of the first-category position points and second estimated coordinates of each of the second-category position points; Determine whether there is a blind area in the indoor space that has not been scanned; If so, returning to the step of scanning the indoor space from any first-category position point in the indoor space; If not, calculating the deviation between the first estimated coordinate and the absolute coordinate; When the deviation is less than a preset threshold, generating a map according to each of the second estimated coordinates; The first estimated coordinates are calculated based on sensor data collected by the robot during movement; The absolute coordinates are coordinates pre-set for known position points. 2. The method according to claim 1, characterized in that the step of determining whether there is a blind area in the indoor space that has not been scanned comprises: Counting the categories to which all position points detected during the scanning of the indoor space belong; Get the preset categories of all location points set for the indoor space; Determine whether the categories to which all detected location points belong completely cover the preset categories; If it is fully covered, it is determined that there is no blind area in the indoor space that has not been scanned; If there is not complete coverage, it is determined that there is a blind area in the room that has not been scanned. 3. A map generation method, applied to a robot, comprising: Starting from any first-category position point in the indoor space, scanning the indoor space; When passing through a first type of position point in the indoor space, obtaining the absolute coordinates corresponding to the first type of position point; Acquire sensor data obtained by scanning the first type of location points and the second type of location points in the indoor space in real time; Determine, based on the sensor data, first estimated coordinates of each of the first-category position points and second estimated coordinates of each of the second-category position points; Determining whether there is a second type of location point in the indoor space that has not been scanned; If so, returning to the step of scanning the indoor space from any first-category position point in the indoor space; If not, calculating the deviation between the first estimated coordinate and the absolute coordinate; When the deviation is less than a preset threshold, a map is generated according to each of the second estimated coordinates. 4. The method according to claim 3, characterized in that the step of determining whether there are any second-type location points in the indoor space that have not been scanned comprises: Count the number of the second type of position points scanned; Get the total number of the second type of position points in the indoor space; The counted number of scanned points is compared with the total number to determine whether there are any second-category position points that have not been scanned in the indoor space. 5. The method according to any one of claims 1 to 4, wherein the first type of location points includes: known location points at the entrance of the lane in the indoor space and/or known location points located in the public area; the second type of location points includes location points corresponding to each storage position on the shelf in the lane in the indoor space and/or location points corresponding to the public area; the method further includes: The first estimated coordinates and the second estimated coordinates are corrected in a closed loop using known position points in the common area. 6. The method according to claim 5, characterized in that the closed-loop correction of the first estimated coordinates and the second estimated coordinates by using known position points in the common area comprises: During the scanning process of the common area, when the robot passes through a known position point in the common area, the estimated coordinates of the known position point in the common area are calculated according to the estimated coordinates of the previous position point and the offset data of the current robot relative to the previous position point; Calculating a second deviation between the estimated coordinates of the known location point of the common area and the corresponding absolute coordinates; When the second deviation is less than a preset threshold, a map is generated according to the second estimated coordinates of the location points corresponding to each of the storage locations. 7. The method according to any one of claims 1 to 4, characterized in that the sensor data includes data collected by multiple sensors; the method further comprises: When the deviation is greater than a preset threshold, adjusting the trust level corresponding to each sensor; Return to the step of scanning the indoor space from any first-category position point in the indoor space; Processing the data collected by the corresponding sensor according to the adjusted confidence level; Based on the processed data, first estimated coordinates of each of the first-category position points and second estimated coordinates of each of the second-category position points are determined. 8. A map generating device, characterized in that the device comprises: A scanning module, used to scan the indoor space from any first type position point in the indoor space; An absolute coordinate acquisition module, used for acquiring the absolute coordinates corresponding to the first type of position points in the indoor space when passing through the first type of position points; A sensor data acquisition module, used for acquiring in real time sensor data obtained by scanning the first type of location points and the second type of location points in the indoor space; an estimated coordinate determination module, configured to determine, based on the sensor data, a first estimated coordinate of each of the first-category position points and a second estimated coordinate of each of the second-category position points; a deviation calculation module, used for calculating the deviation between the first estimated coordinate and the absolute coordinate; A map generation module, configured to generate a map according to each of the second estimated coordinates when the deviation is less than a preset threshold; The scanning module is also used for returning to the step of scanning the indoor space from any first-category position point in the indoor space when there is a blind spot that has not been scanned in the indoor space; and returning to the step of scanning the indoor space from any first-category position point in the indoor space when there is a second-category position point that has not been scanned in the lane; The first estimated coordinates are calculated based on sensor data collected by the robot during movement; the absolute coordinates are coordinates pre-set for known position points. 9. A computer-readable storage medium storing a computer program, wherein when the computer program is executed by a processor, the processor is caused to perform the steps of the method according to any one of claims 1 to 7. 10. A computer device comprising a memory and a processor, wherein the memory stores a computer program, and when the computer program is executed by the processor, the processor executes the steps of the method according to any one of claims 1 to 7.