TWI632346B - A mobile navigation system and method applies nature feature marker and artificial reality. - Google Patents

Abstract

The invention is to improve the conventional indoor navigation technology, to break away from the traditional positioning thinking, and to propose a method and system for indoor navigation using natural feature coding marks and augmented reality. The system is located in the indoor field, at a intersection of all the paths, or at a distance, a navigation positioning point is set, and a natural feature coding mark is arranged at the navigation positioning point, and when the user needs the navigation service, the mobile device can be operated. After the natural feature coded flag is captured, the mobile device decodes and restores, finds the location point of the mobile device in the database, and finds the best path to the navigation destination through the path optimization algorithm, and correspondingly on the path. After navigating the anchor point, the user is guided to move to the next navigation point in augmented reality, and finally guides the user to the destination.

Description

Method and system for indoor navigation using natural feature coding mark and augmented reality mode

The present invention relates to a navigation system, and more particularly to a system for indoor navigation, which utilizes augmented reality and image recognition techniques for positioning and navigation.

Most of the navigation systems of the known technology are positioned by GPS, and then GPS positioning is required to receive the satellite positioning signal. If the satellite positioning signal or the signal is not received in the shaded indoor field, GPS positioning and navigation cannot be performed. In view of this, the conventional technology has a triangular positioning by using radio waves such as Wifi and Bluetooth, and instead of GPS positioning for navigation (such as the Republic of China new patent M463353), it is suitable for general indoor places, but this mode still has to continuously receive wireless. If the signal of the base station is large or the indoor structure is complex, the number of required base stations must be increased. In addition to the cost of the layout, there is still maintenance of the base station and battery replacement. The fingerprint map of the base station is regularly updated. The work has difficulties in subsequent maintenance.

The prior art also has a positioning mark placed in the indoor field, and the mobile device recognizes the positioning mark to obtain the positioning coordinate, and then performs navigation, such as the 2014/8/28 public disclosure. WO2014/128507A2 discloses a method in which a QR Code is used as a navigation mark, and a mobile device scans and recognizes a navigation mark placed in a different place, and compares the navigation mark in the database to obtain a positioning coordinate, and then performs positioning and navigation.

However, this kind of conventional technology also has practical difficulties in the actual indoor field. One is the problem of the accuracy and speed of navigation mark recognition, the other is the directionality of the navigation mark, and the third is that the mark is covered by people or things. Or if some factors are partially damaged, whether the problem can be correctly identified, such as the QR code as a navigation mark by the prior art, the QR code is accurately identified, but the practical time requires 5~10 seconds of identification time, each use Everyone needs to wait 5~10 seconds for each positioning point to be positioned, which is easy to cause the user's intolerance. In addition, the QR code can be identified from different angles, but its identification must rely on the three corners of the back type logo (Finder) Pattern), as the basis for reading the QR code, if a corner is covered by objects such as pedestrians or luggage in the environment, the reading position cannot be recognized, or the mark cannot be recognized, and the QR code is resistant to damage, although Through the coding technology to improve the fault tolerance rate, the current maximum of up to 30%, but on the mobile platform, due to limited mobile phone camera resolution and graphics computing capabilities, the practice can only reach 7%, in other words, the area of the data area is When the coverage is 7%, it cannot be correctly identified, and the information lost by the complex coding technique will also delay the recognition time. In addition to QR code, other 2D barcode recognition technologies, such as Data Matrix Code or MaxiCode, also have similar problems of identification accuracy, speed, directionality or occlusion. There are many people in the indoor navigation field, people move frequently, and the ground signs are covered or partially. Destruction is the normal state of practice, that is, there is a problem that technology cannot be applied.

The object of the present invention is to improve the conventional indoor navigation technology, which is to break away from the traditional positioning thinking, and propose a method for indoor navigation using natural feature coding marks and augmented reality. First, through a specific algorithm, a set of natural feature coding marks with uniqueness, low self-similarity and high spatial frequency is generated, and in all the path intersections in the indoor field to be implemented, or Each time a distance is set, a navigation positioning point is set, and the natural feature coding flag is arranged at the navigation positioning point. Secondly, for all natural feature coding marks, feature extraction is performed through a specific algorithm. The special extraction system converts the natural features of each region in the logo into feature value descriptions through sampling and image segmentation, and finally deposits the natural features. The coded feature value database is used for comparison.

When the user needs the navigation service, the user can select or input the destination on the mobile device, and after the input is completed, the related information is obtained as the pathfinding end information. Secondly, the user scans the natural feature coded flag with the mobile device to extract the feature value of the natural feature code, and the mobile device can compare with the natural feature coded feature value database according to the feature value, and pass the specific algorithm. In the way of statistical inference, find the navigation point that is most likely to correspond. Afterwards, the mobile device is used as the starting point of the pathfinding calculation according to the inference navigation point, and according to the road starting point, the end point and the road network information in the database, the path is optimized, and finally the starting point, the passing navigation point, the entrance and exit, and the ending point are connected. Form a suggested path. The above-mentioned natural feature coding feature value database comparison, finding the starting point of path finding calculation, forming a suggestion path by path optimization mode, etc., in addition to being performed by the mobile device, the mobile device can also obtain the scanned natural feature coded indication. The feature value and the user's pathfinding end point information are transmitted to a cloud server through the network, and the cloud server performs related operations, and then the recommended path is directly transmitted back to the mobile device. And the user enters the destination, And the user scans the natural feature coded flag, and the steps can be interchanged.

After the mobile device obtains the suggested path, it calculates the direction vector of the route between the two points according to the corresponding position of the current navigation point and the next navigation point in the road network, and then calls the electronic compass data of the mobile device itself to find the map side. The orientation of Gebei, and compare this data with the direction vector of the route, calculate the actual azimuth and distance between the navigation point and the next navigation point where the user is currently located, and then project the actual azimuth and distance through the augmented reality mode. On the camera shooting screen, combined with the current map, the user is guided to the next navigation point until the destination is reached. In addition, the user does not have to traverse each navigation point in the process of going to the destination. The user can skip the navigation point in the path at any time according to the needs of the user, and then search for nearby navigation points for scanning when needed, and the system can be dynamically updated. Pathfinding provides a suggested path.

The invention also provides a system for indoor navigation using a natural feature coding mark and an augmented reality mode, the system is used for implementing the above-mentioned natural feature coding mark and augmented reality mode for indoor navigation, which has a unique set. The natural feature coding marks of low self-similarity and high spatial frequency are arranged in the indoor field, at the intersection of all paths, or at the navigation points of each distance. A natural feature coding feature value database stores all feature value descriptions of the natural feature coded flags, and the feature value description is obtained by performing feature extraction by a specific sampling and image segmentation algorithm. And a network database, which stores road network data.

The system further has a mobile device. When the user needs the navigation service, the user can select or input the destination on the mobile device, and after the input is completed, the relevant information is taken out as the pathfinding end information. Secondly, the user scans the natural feature coded flag with the mobile device to obtain the feature value of the natural feature code, and the mobile device can encode the feature value data according to the feature value and the natural feature code. The library performs comparisons and finds the most likely navigation points by statistical inference through a specific algorithm. Afterwards, the mobile device is used as the starting point of the pathfinding calculation according to the inference navigation point, and according to the road starting point, the end point and the road network information in the database, the path is optimized, and finally the starting point, the passing navigation point, the entrance and exit, and the ending point are connected. Form a suggested path. In addition, the system may also have a cloud server, the above-mentioned natural feature database comparison, the starting point of finding the pathfinding algorithm, and the path optimization manner to form a suggestion path, and the feature vector of the natural feature may be encoded by the mobile device. And the user's pathfinding end point information is transmitted to a cloud server through the network, and after the relevant operation is performed by the cloud server, the suggested path is directly transmitted back to the mobile device. In addition, the user inputs the destination, and the user scans the natural feature coded flag, and the steps can be interchanged.

After the mobile device obtains the suggested path, it calculates the direction vector of the route between the two points according to the corresponding position of the current navigation point and the next navigation point in the road network, and then calls the electronic compass data of the mobile device itself to find the map side. The orientation of Gebei, and compare this data with the direction vector of the route, calculate the actual azimuth and distance between the navigation point and the next navigation point where the user is currently located, and then project the actual azimuth and distance through the augmented reality mode. On the camera shooting screen, combined with the current map, the user is guided to the next navigation point until the destination is reached.

In addition, the present invention also proposes a method for generating a natural feature coded mark, which is in a square blank interval, and then randomly generates a high density rectangle, a medium density triangle, a low density line, and then randomly generates white and gray colors. Finally, a unique identification code is generated. If a set of natural feature codes is generated in this way, it is characterized by uniqueness, low self-similarity, and high spatial frequency. The uniqueness means that the natural feature codes are different from each other. The high spatial frequency means that the high-density appearance of the natural features of the code appears on the unit size pattern, and the low self-similarity means that the coded natural feature pattern is taken. One part that needs to be patterned with the natural feature The body similarity is low. When the natural feature coding pattern having the above characteristics is covered, a sufficient and non-self-similar natural feature value can be taken out for the remaining uncovered portions for comparison, with anti-covering, high recognition rate and fast The effect of identification.

201‧‧‧Natural Feature Code Markers

202‧‧‧Mobile devices

251‧‧‧Natural Feature Coded Flag Set

252‧‧‧Mobile devices

253‧‧‧Cloud Server

FIG. 1 is a flow chart of an indoor navigation method according to an embodiment of the present invention.

2A and 2B are diagrams showing the architecture of an indoor navigation mode according to an embodiment of the present invention.

FIG. 3A is a schematic diagram showing the generation of a natural feature coded flag.

FIG. 3B is a schematic diagram showing a feature point of obtaining a natural feature code by using sampling.

FIG. 3C is a schematic diagram showing the transformation of a region natural feature into a feature vector by a pattern segmentation.

FIG. 4A is a schematic diagram of a user selecting or inputting a destination.

FIG. 4B is a schematic diagram of a user scanning a natural feature coded flag.

FIG. 5A illustrates a schematic diagram of guiding to the next navigation point in an augmented reality manner.

FIG. 5B is a schematic diagram showing another way of guiding to the next navigation point in augmented reality.

Referring to FIG. 1, there is shown a flow chart of a method for indoor navigation according to an embodiment of the present invention. Firstly, a method is used to generate a set of natural feature coding marks with uniqueness, low self-similarity and high spatial frequency. The natural feature coding system is a human-like recognition principle and has no standard format coding method. It replaces the coding modes of 0 and 1, black and white in the traditional two-dimensional bar code with natural features, and replaces the decoding by statistical inference. An embodiment of the invention The natural feature coding used in the method is generated by a special method, which is in a square blank interval, and then randomly generates high-density rectangles, medium-density triangles, low-density lines, and then randomly generates white and gray colors. Finally, a unique identification code is generated (see Figure 3A). In this method, a set of natural feature codes is generated, which can have the characteristics of uniqueness, low self-similarity and high spatial frequency.

The uniqueness means that the natural feature codes are different from each other. The high spatial frequency means that the high-density appearance of the natural features of the code appears on the unit size pattern, and the low self-similarity means that the coded natural feature pattern is taken. One part needs to be similar to the natural feature pattern itself. Therefore, when the natural feature coding pattern having the above characteristics is covered, it is still possible to take out sufficient and non-self-similar natural feature values for comparison in the remaining uncovered portions, so that the natural feature coding is more conventional. 2D barcode or QR code with anti-covering, high recognition rate and quick identification.

After the set of natural feature coding flags is generated, the flag is extracted by the specific algorithm through the specific algorithm, and one or a set of feature values are generated for each natural feature code, and stored in the natural feature coded feature value database. In an embodiment of the present invention, the feature value information acquisition is performed by first sampling and sampling comparison, and converting a graphic of a natural feature signature into a pixel point, and the pixel difference between the pixel and the adjacent point is up to one. The point of the critical value is defined as the feature point (see Fig. 3B). Then, after the image is segmented into different regions, the center of gravity of the feature point distribution is calculated according to the distribution of the feature points of each region, and The vector relationship between the center of gravity and the origin of the region is used as the feature vector (see Fig. 3C). Since the vector has a scale-invariant characteristic, it can resist the effects of light, noise, viewing angle and shadowing. After the feature vector of each region, it is used as a set of feature values of the natural feature coding and is stored in the database.

At the same time, in the indoor field of the desired implementation, select a route intersection, or set a navigation point at intervals. In an embodiment of the invention, the navigation points can be viewed from each other, so that the user can visually assist the navigation. The natural feature coded mark generated in the previous step is fixed on the navigation point, and the fixed position can be on the floor, the ceiling, or the three-dimensional wall surface or the fixed object.

When the navigation is to be located, the destination information of the user to reach the destination is first obtained. In an embodiment, the destination information may be selected by the user or directly input (see FIG. 4A). In another embodiment, After obtaining the destination information, it can be compared with the road network database to extract relevant information such as coordinates and floor, as the pathfinding end information. And the mobile device scans the natural feature coded flag in its vicinity. In an embodiment, the step can be guided by the mobile device to the user (see FIG. 4B) and can be scanned by the mobile device front or rear camera. After scanning the natural feature coded flag, the feature value information of the flag is captured by the same algorithm to obtain a set of feature values of the flag. In an embodiment, the feature value is a set of feature vectors.

Then, according to the feature values obtained by the scan and the feature values stored in the natural feature coded feature value database, through a specific algorithm, the most likely corresponding navigation point is found by way of statistical inference, as a mobile device. Pathfinding starting point information. In an embodiment, the method of statistical inference can perform the inner product of the feature vector obtained by scanning and the feature vector in the database, and find the result of the maximum value of the operation result, which is the most likely corresponding Navigation point.

Then, based on the starting point information, the end point information and the road network database data, the path finding algorithm is used for path finding calculation. In an embodiment, the pathfinding calculation may be a three-degree spatial pathfinding calculation algorithm, which first searches for the minimum floor to pass through, and then searches for the optimal entrance and exit corresponding to each floor. Finally, the starting point, the passing navigation point, the entrance and exit, and the ending point are formed to form a suggested path.

The mobile device extracts the information of the next navigation point at the location according to the navigation point information in the suggested path, and calculates the direction vector of the road network according to the road network information of the current navigation point and the next navigation point. And through the electronic compass of the mobile device, the orientation of the mobile device is detected. Finally, through the direction vector of the road network and the orientation of the user, calculate the actual azimuth and distance between the mobile device and the next navigation point, and project the actual azimuth and distance onto the camera screen through the augmented reality mode. In combination with the present scene, in one embodiment, the seeker icon can be projected onto the camera screen in combination with the current scene (see Figure 5A). In another embodiment, the mobile device is projected onto the camera screen through a virtual marking or guide line and combined with the current scene to indicate the direction and distance of travel to the next positioning point (see FIG. 5B).

Thereby, the user can go to the next navigation point according to the instruction of the marker, and when the next navigation point is reached, the mobile device scans the natural feature code editor again, and guides the user to the next step in the above manner. Navigate the point until you reach your destination. In an embodiment, the user does not have to traverse each navigation point in the process, and the user can skip the navigation point in the path at any time according to the needs of the user, and if necessary, look for a nearby navigation point to scan at any time, and the system can dynamically search for the time. The road provides a suggested path.

The steps of comparing the above-mentioned natural feature coded flag database, finding the starting point of the path finding calculus, and forming the suggested path by the path optimization mode may be performed by the mobile device in one embodiment, and may be performed by the mobile device in another embodiment. The natural feature coded flag feature value is uploaded to the cloud server by the network, and after the step is performed by the cloud server, the suggested path is transmitted back to the mobile device.

Please refer to FIG. 2A, which is a schematic diagram of a system for indoor navigation according to an embodiment of the present invention. The system is used to implement the indoor navigation method of the present invention, which has a set of There is a unique feature, low self-similarity and high spatial frequency natural feature coding mark (201), which is placed in the indoor field, at the intersection of the path, or at a navigation point of each distance. A natural feature coding feature value database stores all self-characteristic coded value descriptions, and the feature value description is obtained by performing feature extraction by a specific sampling and image segmentation algorithm. And a network database, which stores road network data.

The system further has a mobile device having at least one input unit, a front or rear camera, an arithmetic unit, a storage unit, an electronic compass, and a display unit. The storage unit of the mobile device stores the above-mentioned natural feature coded feature value database and the road network database. When the user wants to perform navigation and positioning, the input unit first inputs the destination information of the user to reach the destination. In an embodiment, the destination information may be selected by the user or directly input, and in another implementation. In the example, after obtaining the destination information, it can be compared with the road network database, and relevant information such as coordinates and floor is taken out as the pathfinding end information. The native feature coded flag is scanned by the mobile device's front or rear camera. In one embodiment, the scan can be directed by the mobile device to the user. Then, the arithmetic unit, after performing the algorithm of the specific sampling and the image segmentation, encodes the feature value information of the flag after the natural feature is encoded to obtain a set of feature values of the flag. In an embodiment, the The feature value is a set of feature vectors.

Then, the arithmetic unit compares the feature values obtained by the scan with the feature values stored in the natural feature coded feature value database in the storage unit, and compares them by statistical inference through a specific algorithm to find the most likely corresponding The navigation point is used as a starting point for the mobile device. In an embodiment, the statistical inference method can perform the inner product product of the feature vector obtained by scanning and the feature vector in the database, and find the knot of the maximum value of the operation result. As a navigation point that is most likely to correspond.

Then, the arithmetic unit performs path finding calculation through the path finding algorithm according to the starting point information, the end point information, and the road network database data. In an embodiment, the pathfinding calculation may be a three-degree spatial pathfinding calculation algorithm, which first searches for the minimum floor to pass through, and then searches for the optimal entrance and exit corresponding to each floor. Finally, the starting point, the passing navigation point, the entrance and exit, and the ending point are formed to form a suggested path. The relevant data is then stored in the storage unit.

Then, the computing unit extracts the information of the next navigation point at the location according to the navigation point information in the suggested path, and calculates the direction vector of the road network according to the road network information of the current navigation point and the next navigation point. And by the electronic compass of the mobile device, the orientation of the mobile device is detected. Finally, through the direction vector of the road network and the orientation of the user, calculate the actual azimuth and distance between the mobile device and the next navigation point, and project the actual azimuth and distance onto the camera screen through the augmented reality mode. In combination with the present scene, in one embodiment, the seeker icon can be projected onto the camera screen in combination with the current scene (see Figure 5A). In another embodiment, the mobile device is projected onto the camera screen through a virtual marking or guide line and combined with the current scene to indicate the direction and distance of travel to the next positioning point (see FIG. 5B).

In this way, the user can go to the next navigation point according to the instruction of the marker, and when the next navigation point is reached, the mobile device scans the natural feature code editor again, and guides the user to the next navigation point in the above manner. Until you reach your destination. In an embodiment, the user does not have to traverse each navigation point in the process, and the user can skip the navigation point in the path at any time according to the needs of the user, and if necessary, look for a nearby navigation point to scan at any time, and the system can dynamically search for the time. The road provides a suggested path.

2B is a schematic diagram showing a system for indoor navigation according to another embodiment of the present invention. In the embodiment, the cloud server has a cloud server, a computing unit, and a storage unit, and the mobile device has another network unit. In addition, the natural feature encoding feature value database and the road network database are stored in the cloud server.

When the user wants to perform navigation and positioning, the mobile device obtains the pathfinding end point information of the user's destination and the characteristic value of the natural feature coded flag in the vicinity of the user, and then transmits the above two information through the network unit of the mobile device. The result is transmitted to the cloud server, and then the computing unit of the cloud server compares with the natural feature coded flag database to find the starting point of the pathfinding calculation, and the recommended path is calculated by the path optimization mode, and then the cloud server is passed through the cloud server. The network unit transmits the suggested path back to the mobile device and is stored in the storage unit of the mobile device. And the operation unit of the mobile device continues the projection processing of the subsequent augmented reality, guiding the user to the next navigation point for navigation.

The above is only an embodiment of the present invention, and is not intended to limit the scope of the patent protection of the present invention. Any changes, refinements, and equivalent substitutions made by those skilled in the art without departing from the spirit and scope of the invention are still within the scope of the invention.

Claims (10)

An indoor navigation method, characterized in that a natural feature coding flag is used and navigation is performed in an augmented manner, the method comprising the steps of: generating a set of natural feature coded marks; the set of natural feature coded marks is generated in a manner In the blank base map of size, the high-density rectangle, the medium-density triangle, and the low-density line are randomly generated, and then the colors of white and gray are randomly generated; the natural feature coded flag is arranged on the navigation point of the indoor field; Extracting the feature value of the natural feature coded flag and storing it in the natural feature coded feature value database; obtaining the end point information of the navigation destination; scanning the natural feature coded mark near the navigation start point by the mobile device, and taking the scan to obtain The feature value of the natural feature coded flag; the feature value of the natural feature coded flag obtained by the scan is compared with the feature value in the natural feature coded feature value database to find the navigation start point that is most likely to correspond; according to the navigation start point , navigation destinations and road network information, path optimization, to obtain suggested paths; The moving device calculates the direction vector of the route between the two points according to the information of the navigation starting point and the next navigation positioning point in the suggested path; and calls the electronic compass to obtain the current orientation of the mobile device, and finally converts the direction vector into the actual orientation and distance; The actual orientation and distance are displayed on the camera shooting screen, and the navigation is guided to the next navigation positioning point by augmented reality. The method of claim 1, wherein the step of scanning the extracted natural feature coding feature value, comparing the feature value in the natural feature coding feature value database, and the path optimization step may be performed by the action After the device downloads and stores the natural feature coded feature value database and the road network information, the recommended path is obtained by performing an offline operation, and the mobile device extracts the natural feature coded feature value of the scan to a cloud server. After the eigenvalue comparison and the path optimization step are performed by the cloud server, the recommended path is obtained, and then the suggested path is transmitted back to the mobile device. For example, the method of claim 2, wherein the actual orientation is displayed on the camera shooting screen, can be displayed in the manner of the marker display, or displayed as a virtual marking or guiding line. The method of any one of claims 1 to 3, wherein the step of extracting the feature value of the natural feature coded flag is performed by: converting the natural feature coded flag into a pixel and converting it into a pixel; Whether the difference between each pixel point and its neighboring pixel point reaches a threshold value, if it is reached, the pixel point is marked as a feature point; the natural feature coding flag pattern is divided into several regions, and the distribution of the feature points according to each region And the relative relationship with the origin of the region, transforming the feature point distribution into a feature vector of the region; After each region is calculated one by one according to the previous steps, a set of feature vectors is obtained, which can be used as the feature value of the natural feature coding flag. For example, in the method of claim 4, wherein the natural feature coded flag is generated, a unique identification code may be generated after the foregoing steps. A system for indoor navigation, characterized in that a natural feature coding flag is used and navigation is performed in an augmented manner, the system comprising: a set of natural feature coding marks arranged on the navigation positioning point; and a generation mode of the natural feature coding flag , in a blank base map of considerable size, successively randomly generate high-density rectangles, medium-density triangles, low-density lines, and then randomly generate white and gray colors; a natural feature coded feature value database, the natural features The coded feature value database stores all the self-characteristic coded value descriptions; the one-way network database stores the road network data; and the mobile device has at least one input unit, one front or one rear camera, and one arithmetic unit a storage unit, an electronic compass, and a display unit; the input unit is configured to obtain end point information of the navigation destination; the front or rear camera is configured to scan a natural feature coded mark near the navigation start point; The unit is used to capture the feature value of the natural feature coded mark obtained by the camera scanning; a natural feature coded flag feature value comparison module, which compares the feature value of the natural feature coded mark obtained by the scan with the feature value in the natural feature coded feature value database to find the most likely corresponding a navigation starting point; a path optimization module, which performs path optimization according to the navigation starting point, the navigation destination, and the road network database to obtain a suggested path; the computing unit of the mobile device is also used according to the navigation starting point and Suggesting the information of the next navigation anchor point in the path, calculating the direction vector of the route between the two points; and calling the electronic compass to obtain the current orientation of the mobile device, and finally converting the direction vector into the actual orientation and distance; The device displays the actual orientation and distance together with the picture taken by the front or rear camera on the display unit, and guides the navigation to the next navigation point by augmented reality. The system of claim 6, wherein the natural feature coded flag feature value comparison module and the path optimization module are included in an operation unit of the mobile device, and are stored by the mobile unit of the mobile device. After the natural feature encodes the feature value database and the road network database, the operation path is obtained by the operation unit to obtain the suggested path; and the natural feature coded flag feature value comparison module and the path optimization module are also The cloud server has a storage unit for storing the natural feature coded feature value database and the network database, and a network unit; The mobile device further has a network unit, and the eigenvalue ratio of the computing unit of the cloud server is performed by transmitting the natural feature coded feature value obtained by the scan to the cloud server through the other network unit. After the path is optimized, the recommended path is obtained, and then the recommended path is transmitted back to the mobile device through the network unit of the cloud server. The system of claim 7, wherein the manner in which the actual orientation is displayed on the display unit can be displayed in the manner of a marker, or displayed as a virtual marking or a guide line. The system of any one of claims 6 to 8, wherein the step of extracting the feature value of the natural feature coded flag is performed by: converting the natural feature coded flag into a pixel and converting it into a pixel; Whether the difference between each pixel point and its neighboring pixel point reaches a threshold value, if it is reached, the pixel point is marked as a feature point; the natural feature coding flag pattern is divided into several regions, and the distribution of the feature points according to each region And the relative relationship with the origin of the region, the feature point distribution is transformed into the feature vector of the region; each region is calculated one by one according to the previous steps, and a set of feature vectors is obtained, which can be used as the feature of the natural feature coding flag. value. For example, in the system of claim 9th, wherein the natural feature coded flag is generated, a unique identification code may be generated after the foregoing steps.