Position identification method and device, computer equipment and storage medium

Abstract

The invention relates to a position identification method and device, computer equipment, a storage medium and a computer program product, and can be applied to the technical field of computers. The method comprises the steps that the current position of a to-be-recognized object is recognized from an obtained to-be-recognized image; when the current position coincides with the target position of the to-be-recognized object, intercepting a region image corresponding to the to-be-recognized object from the to-be-recognized image; performing vertex position identification on the to-be-identified object in the regional image through a vertex position identification model to obtain a vertex position of the to-be-identified object; recognizing the vertex position and the target position to obtain a position recognition result of the to-be-recognized object; the position recognition result is used for representing whether the to-be-recognized object is located in the target position. By adopting the method, the efficiency of position identification can be improved.

Description

Technical Field

[0001] This application relates to the field of computer technology, and in particular, to a position recognition method, apparatus, computer device, storage medium, and computer program product. Background Art

[0002] With the development of intelligent technologies, position recognition has important applications in many fields. By recognizing the positions of objects, their exact locations in space can be understood, which is crucial for resource optimization, path planning, inventory management, etc. Therefore, how to perform position recognition efficiently has become an important research direction.

[0003] Traditional technologies usually perform position recognition by means of manual visual observation; however, performing position recognition in this way requires a lot of manual processing time, resulting in low efficiency of position recognition. Summary of the Invention

[0004] Based on this, in view of the above technical problems, it is necessary to provide a position recognition method, apparatus, computer device, computer-readable storage medium, and computer program product that can improve the efficiency of position recognition.

[0005] In a first aspect, this application provides a position recognition method. The method includes:

[0006] Recognize the current position of the object to be recognized from the acquired image to be recognized;

[0007] When there is an overlap between the current position and the target position of the object to be recognized, extract the region image corresponding to the object to be recognized from the image to be recognized;

[0008] Through the vertex position recognition model, perform vertex position recognition on the object to be recognized in the region image to obtain the vertex position of the object to be recognized;

[0009] Perform recognition on the vertex position and the target position to obtain a position recognition result of the object to be recognized; the position recognition result is used to indicate whether the object to be recognized is located within the target position.

[0010] In one embodiment, the performing recognition on the vertex position and the target position to obtain a position recognition result of the object to be recognized includes:

[0011] Perform recognition on the vertex position and the target position to obtain a vertex position recognition result of the object to be recognized; the vertex position recognition result is used to indicate whether the vertices of the object to be recognized are located within the target position;

[0012] Determine the current number of the vertices located within the target position according to the vertex position recognition result;

[0013] In the case where the current number is greater than or equal to a preset number, confirm that the position recognition result of the object to be recognized indicates that the object to be recognized is located within the target position.

[0014] In one embodiment, recognizing the current position of the object to be recognized from the acquired image to be recognized includes:

[0015] Acquire an image to be recognized; the image to be recognized includes the object to be recognized;

[0016] Perform position recognition on the object to be recognized in the image to be recognized through an object position recognition model to obtain the current position of the object to be recognized.

[0017] In one embodiment, in the case where there is an overlap between the current position and the target position of the object to be recognized, cropping out the region image corresponding to the object to be recognized from the image to be recognized includes:

[0018] In the case where there is an overlap between the current position and the candidate position of the object to be recognized, crop out the region image from the image to be recognized;

[0019] Use the candidate position that overlaps with the current position as the target position.

[0020] In one embodiment, the method further includes:

[0021] Recognize a candidate region from the image to be recognized;

[0022] Determine the position of the candidate region as the candidate position of the object to be recognized.

[0023] In one embodiment, after obtaining the position recognition result of the object to be recognized by recognizing the vertex position and the target position, the method further includes:

[0024] Obtain the first historical position recognition result corresponding to the first recognized image; the first recognized image is an image taken within a first preset time period before the shooting time of the image to be recognized;

[0025] In the case where the first historical position recognition result and the position recognition result are both consistent, obtain the second historical position recognition result corresponding to the second recognized image; the second recognized image is an image taken within a second preset time period before the shooting time of the image to be recognized;

[0026] Determine the object state recognition result corresponding to the target position according to the position recognition result, the first historical position recognition result, and the second historical position recognition result; the object state recognition result is used to indicate whether the object to be recognized is in the target position.

[0027] In one embodiment, after determining the object state recognition result corresponding to the target position, the method further includes:

[0028] In the case where the object state recognition result indicates that the object to be recognized is in the target position, generate a transportation instruction corresponding to the object to be recognized;

[0029] Send the transportation instruction to a transportation device; the transportation device is used to transport the object to be recognized from the target position to the transportation position of the object to be recognized.

[0030] In a second aspect, the present application further provides a position recognition device. The device includes:

[0031] An object recognition module, configured to recognize the current position of an object to be recognized from the acquired image to be recognized;

[0032] An image cropping module, configured to crop a region image corresponding to the object to be recognized from the image to be recognized when there is an overlap between the current position and the target position of the object to be recognized;

[0033] A vertex recognition module, configured to perform vertex position recognition on the object to be recognized in the region image through a vertex position recognition model to obtain the vertex position of the object to be recognized;

[0034] A target recognition module, configured to recognize the vertex position and the target position to obtain a position recognition result of the object to be recognized; the position recognition result is used to indicate whether the object to be recognized is in the target position.

[0035] In a third aspect, the present application further provides a computer device. The computer device includes a memory and a processor, the memory stores a computer program, and when the processor executes the computer program, the following steps are implemented:

[0036] Recognize the current position of the object to be recognized from the acquired image to be recognized;

[0037] When there is an overlap between the current position and the target position of the object to be recognized, crop a region image corresponding to the object to be recognized from the image to be recognized;

[0038] Using the vertex position recognition model, recognize the vertex positions of the object to be recognized in the region image, and obtain the vertex positions of the object to be recognized;

[0039] Recognize the vertex positions and the target positions to obtain a position recognition result of the object to be recognized; the position recognition result is used to indicate whether the object to be recognized is within the target position.

[0040] In a fourth aspect, the present application also provides a computer-readable storage medium. The computer-readable storage medium has a computer program stored thereon, and when the computer program is executed by a processor, the following steps are implemented:

[0041] Recognize the current position of the object to be recognized from the acquired image to be recognized;

[0042] When there is an overlap between the current position and the target position of the object to be recognized, extract the region image corresponding to the object to be recognized from the image to be recognized;

[0043] Using the vertex position recognition model, recognize the vertex positions of the object to be recognized in the region image, and obtain the vertex positions of the object to be recognized;

[0044] Recognize the vertex positions and the target positions to obtain a position recognition result of the object to be recognized; the position recognition result is used to indicate whether the object to be recognized is within the target position.

[0045] In a fifth aspect, the present application also provides a computer program product. The computer program product includes a computer program, and when the computer program is executed by a processor, the following steps are implemented:

[0046] Recognize the current position of the object to be recognized from the acquired image to be recognized;

[0047] When there is an overlap between the current position and the target position of the object to be recognized, extract the region image corresponding to the object to be recognized from the image to be recognized;

[0048] Using the vertex position recognition model, recognize the vertex positions of the object to be recognized in the region image, and obtain the vertex positions of the object to be recognized;

[0049] Recognize the vertex positions and the target positions to obtain a position recognition result of the object to be recognized; the position recognition result is used to indicate whether the object to be recognized is within the target position.

[0050] The above-mentioned position recognition method, device, computer device, storage medium and computer program product identify the current position of the object to be recognized from the acquired image to be recognized; when there is an overlap between the current position and the target position of the object to be recognized, extract the region image corresponding to the object to be recognized from the image to be recognized; through the vertex position recognition model, recognize the vertex position of the object to be recognized in the region image to obtain the vertex position of the object to be recognized; recognize the vertex position and the target position to obtain the position recognition result of the object to be recognized; the position recognition result is used to indicate whether the object to be recognized is within the target position. This solution identifies the current position of the object to be recognized from the image to be recognized. When there is an overlap between the current position and the target position of the object to be recognized, extract the region image corresponding to the object to be recognized, recognize the vertex position of the object to be recognized in the region image through the vertex position recognition model, and recognize the vertex position and the target position to obtain the position recognition result indicating whether the object to be recognized is within the target position. In this way, this solution determines the position state of the object to be recognized by recognizing the preliminary position and the precise position of the object to be recognized and comparing them with the target position, which is beneficial to improving the efficiency and accuracy of position recognition. BRIEF DESCRIPTION OF THE DRAWINGS

[0051] In order to more clearly illustrate the technical solutions in the embodiments of the present application or related technologies, the following will briefly introduce the drawings required for use in the description of the embodiments or related technologies. Obviously, the following drawings are only some embodiments of the present application. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0052] Figure 1 It is a schematic flowchart of the position recognition method in an embodiment;

[0053] Figure 2 It is a schematic flowchart of the steps for determining the position recognition result in an embodiment;

[0054] Figure 3 It is a schematic flowchart of the steps for recognizing the current position in an embodiment;

[0055] Figure 4 It is a schematic flowchart of the steps for determining the target position in an embodiment;

[0056] Figure 5 It is a schematic flowchart of the steps for determining the candidate position in an embodiment;

[0057] Figure 6 It is a structural block diagram of the position recognition device in an embodiment;

[0058] Figure 7Internal structure diagram of a computer device in an embodiment. Detailed implementation manners

[0059] In order to make the objectives, technical solutions and advantages of the present application more clear and understandable, the present application will be further described in detail below with reference to 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.

[0060] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data for analysis, stored data, displayed data, etc.) involved in the present application are all information and data authorized by the user or fully authorized by all parties, and the collection, use and processing of relevant data need to comply with relevant regulations.

[0061] At present, cage trucks are used for cargo transfer in the transfer yard. Due to their certain weight, the transfer efficiency in the yard is not high. Therefore, AGVs (Automated Guided Vehicles) can be introduced to transfer cage trucks in the yard; tasks are issued in the task management system of the site, and then the AGVs transport the specified cage trucks from the specified starting point to the specified ending point to complete these tasks. For tasks with known starting points and known ending points, the task issuing process of the AGV can be manual or automatic. The present application provides a set of systems for automatically calling AGVs to transfer cage trucks, and the key points are as follows: 1. Cage truck identification method; 2. Cage truck key point positioning method; 3. Cage truck in-place detection method; 4. Automatic calling system.

[0062] Based on this, the present application provides a position recognition method, device, computer device, storage medium and computer program product. First, the position recognition method provided by the present application will be described.

[0063] In an exemplary embodiment, as Figure 1 shown, a position recognition method is provided. In this embodiment, an example is given where this method is applied to a terminal; it can be understood that this method can also be applied to a server, and can also be applied to a system including a terminal and a server, and is implemented through the interaction between the terminal and the server. Among them, the terminal can be but is not limited to various personal computers, laptop computers, smart phones, tablet computers, etc.; the server can be implemented by an independent server or a server cluster composed of multiple servers. In this embodiment, the method includes the following steps:

[0064] Step S101, identify the current position of the object to be recognized from the acquired image to be recognized.

[0065] Among them, the image to be recognized can be the original image data that needs to be position-recognized acquired through devices such as cameras.

[0066] Among them, the object to be recognized can be an object whose position needs to be recognized. For example, the object to be recognized can be a cage car. For instance, in a logistics scenario, the object to be recognized can be a cage car applied to the logistics scenario.

[0067] Among them, the current position can be the initial position coordinates of the object to be recognized recognized in the image to be recognized by a position recognition model.

[0068] Optionally, the terminal collects on-site image data in real time through devices such as a camera to obtain the image to be recognized; uses a position recognition model to recognize the current position coordinates of the object to be recognized (such as a cage car) in the image to be recognized, and obtains the current position of the object to be recognized.

[0069] Step S102, when there is an overlap between the current position and the target position of the object to be recognized, extract the region image corresponding to the object to be recognized from the image to be recognized.

[0070] Among them, the target position can be the end position that the object to be recognized needs to reach. For example, the target position can be the position of a storage location.

[0071] Among them, the region image can be a sub-image corresponding to the object to be recognized extracted from the image to be recognized.

[0072] Optionally, the terminal determines whether there is an overlap between the current position of the object to be recognized and the target position of the object to be recognized (such as a storage location). If there is an overlap between the current position and the target position, then extract the image region corresponding to the object to be recognized from the image to be recognized according to the coordinates of the current position, and obtain the region image corresponding to the object to be recognized.

[0073] Step S103, through a vertex position recognition model, perform vertex position recognition on the object to be recognized in the region image to obtain the vertex position of the object to be recognized.

[0074] Among them, the vertex position recognition model can be a deep learning model used to recognize the specific contour or structural feature point positions of the object to be recognized in the region image. For example, the vertex position recognition model is used to recognize the four vertex coordinates of a cage car.

[0075] Among them, the vertex position can be the feature point coordinates of the object to be recognized recognized by the vertex position recognition model. For example, the four vertex coordinates of the object to be recognized.

[0076] Optionally, the terminal inputs the regional image into the vertex position recognition model, and through the vertex position recognition model, recognizes the vertex position of the object to be recognized in the regional image, and obtains the vertex position of the object to be recognized output by the vertex position recognition model, such as the specific coordinates of the four vertices of the object to be recognized (cage car).

[0077] Step S104: Recognize the vertex position and the target position to obtain the position recognition result of the object to be recognized; the position recognition result is used to indicate whether the object to be recognized is within the target position.

[0078] Among them, the position recognition result can be the recognition result of judging whether the object to be recognized is completely within the target position according to the comparison relationship between the vertex position and the target position.

[0079] Optionally, the terminal compares the recognized vertex position with the target position (such as the storage location). If at least three points in the vertex position are within the target position range, it is determined that the object to be recognized is already within the target position; otherwise, it is determined that the object to be recognized is not within the target position; according to whether the object to be recognized is within the target position, generate and output the position recognition result of the object to be recognized.

[0080] In the above position recognition method, from the obtained image to be recognized, recognize the current position of the object to be recognized; in the case where there is an overlap between the current position and the target position of the object to be recognized, intercept the regional image corresponding to the object to be recognized from the image to be recognized; through the vertex position recognition model, recognize the vertex position of the object to be recognized in the regional image to obtain the vertex position of the object to be recognized; recognize the vertex position and the target position to obtain the position recognition result of the object to be recognized; the position recognition result is used to indicate whether the object to be recognized is within the target position. This solution recognizes the current position of the object to be recognized from the image to be recognized. When there is an overlap between the current position and the target position of the object to be recognized, intercept the regional image corresponding to the object to be recognized, recognize the vertex position of the object to be recognized in the regional image through the vertex position recognition model, and recognize the vertex position and the target position to obtain the position recognition result of whether the object to be recognized is within the target position; in this way, this solution recognizes the preliminary position and the precise position of the object to be recognized and compares them with the target position, thereby judging the position state of the object to be recognized, which is beneficial to improving the efficiency and accuracy of position recognition.

[0081] In an exemplary embodiment, as Figure 2 shown, in step S104, recognizing the vertex position and the target position to obtain the position recognition result of the object to be recognized specifically includes the following content:

[0082] Step S201: Identify the vertex positions and the target position to obtain the vertex position recognition result of the object to be recognized; the vertex position recognition result is used to indicate whether the vertices of the object to be recognized are within the target position.

[0083] Step S202: Determine the current number of vertices located within the target position according to the vertex position recognition result.

[0084] Step S203: When the current number is greater than or equal to the preset number, confirm that the position recognition result of the object to be recognized indicates that the object to be recognized is within the target position.

[0085] Among them, the vertex position recognition result can be the recognition result obtained by comparing with the target position coordinates to determine whether each vertex of the object to be recognized is within the target position range.

[0086] Among them, the current number can be the current number of vertices that have been confirmed to be within the target position statistically obtained from the vertex position recognition result.

[0087] Among them, the preset number can be a set quantity threshold for determining whether the object to be recognized has completely entered the target position.

[0088] Optionally, the terminal compares the recognized vertex position coordinates with the target position coordinates to determine whether each vertex is within the target position range; according to the comparison result between the vertex position and the target position, obtain the vertex position recognition result of each vertex of the object to be recognized, and this result indicates whether the vertex is within the target position; according to the obtained vertex position recognition results of all vertices, count the current number of vertices that have been confirmed to be within the target position; compare the current number with the preset number, where the preset number can be the minimum number of vertices that can be used to determine that the object to be recognized has completely entered the target position set in advance; if the current number is greater than or equal to the preset number, it indicates that a sufficient number of vertices of the object to be recognized have entered the target position, indicating that the object to be recognized is within the target position, and then confirm that the position recognition result of the object to be recognized indicates that the object to be recognized is within the target position.

[0089] The technical solution provided in this embodiment further determines whether the whole object to be recognized has entered the target position by counting the number of vertices entering the target position, and obtains a more accurate position recognition result, which is beneficial to improving the accuracy of position recognition.

[0090] In an exemplary embodiment, as Figure 3 shown, in step S101, from the obtained image to be recognized, identify the current position of the object to be recognized, which specifically includes the following content:

[0091] Step S301: Obtain the image to be recognized; the image to be recognized includes the object to be recognized.

[0092] Step S302: Use the object position recognition model to recognize the position of the object to be recognized in the image to be recognized, and obtain the current position of the object to be recognized.

[0093] Among them, the object position recognition model can be a neural network model for recognizing the position of the object to be recognized.

[0094] Optionally, the terminal obtains the image to be recognized. For example, it obtains the original image data containing the object to be recognized; inputs the obtained image to be recognized into the pre-trained object position recognition model, and uses the pre-trained object position recognition model to recognize the position of the object to be recognized in the image to be recognized, and obtains the current position of the object to be recognized output by the pre-trained object position recognition model.

[0095] For example, the terminal inputs the obtained image to be recognized into the pre-trained object position recognition model, analyzes and calculates the image to be recognized through the pre-trained object position recognition model, extracts the image features of the image to be recognized, recognizes the image features, locates the position of the object to be recognized in the image to be recognized, and obtains the current position information of the object to be recognized. Among them, the current position information can include the abscissa and ordinate of the object to be recognized in the image coordinate system.

[0096] The technical solution provided in this embodiment processes the image through the object position recognition model, and then locates the current position of the object to be recognized in the image, realizing the process of automatically recognizing the object position, which is beneficial to improving the efficiency and accuracy of position recognition.

[0097] In an exemplary embodiment, as Figure 4 shown, in step S102, when there is an overlap between the current position and the target position of the object to be recognized, the region image corresponding to the object to be recognized is intercepted from the image to be recognized, which specifically includes the following content:

[0098] Step S401: When there is an overlap between the current position and the candidate position of the object to be recognized, intercept the region image from the image to be recognized;

[0099] Step S402: Use the candidate position that overlaps with the current position as the target position.

[0100] Among them, the candidate position can be the candidate position that the object to be recognized can reach. For example, the candidate position can be the position coordinate range of different storage locations.

[0101] Optionally, the terminal determines whether there is an overlap between the current position and the candidate positions of the object to be recognized; in the case where there is an overlap between the current position and the candidate positions, a regional image is cropped from the image to be recognized; the candidate position that overlaps with the current position is used as the target position of the object to be recognized.

[0102] The technical solution provided in this embodiment automatically extracts a regional image from the image to be recognized according to the overlap relationship between the current position and the candidate positions, and identifies the target position of the object to be recognized, which is conducive to accurately determining the regional image and the target position, thereby improving the accuracy of position recognition.

[0103] In an exemplary embodiment, as Figure 5 shown, it further includes the step of determining candidate positions, which specifically includes the following:

[0104] Step S501: Identify a candidate region from the image to be recognized;

[0105] Step S502: Determine the position of the candidate region as the candidate position of the object to be recognized.

[0106] Among them, the candidate region may be a candidate arrival region of the object to be recognized included in the image to be recognized. For example, the candidate region may be the region of each storage location.

[0107] Among them, the position of the candidate region may be the coordinate point or the coordinate range of the candidate region.

[0108] Optionally, the terminal identifies a candidate region containing the object to be recognized from the image to be recognized; locates each recognized candidate region in the image coordinate system to determine its position coordinates; uses the position coordinates of each candidate region as the candidate position set of the object to be recognized, thus completing the process of recognizing the candidate region from the original image and determining its position as the candidate position.

[0109] The technical solution provided in this embodiment automatically identifies the candidate region in the image to be recognized and determines its position as the candidate position of the object to be recognized, which is conducive to accurately determining the candidate position, thereby improving the accuracy of position recognition.

[0110] In an exemplary embodiment, after identifying the vertex position and the target position to obtain the position recognition result of the object to be recognized, the following is further included: obtaining a first historical position recognition result corresponding to a first recognized image; the first recognized image is an image taken within a first preset time period before the shooting time of the image to be recognized; in the case where the first historical position recognition result and the position recognition result are both consistent, obtaining a second historical position recognition result corresponding to a second recognized image; the second recognized image is an image taken within a second preset time period before the shooting time of the image to be recognized; determining an object state recognition result corresponding to the target position according to the position recognition result, the first historical position recognition result, and the second historical position recognition result; the object state recognition result is used to indicate whether there is an object to be recognized within the target position.

[0111] Among them, the first recognized image may be a series of historical images taken within a first preset time period before the shooting time of the image to be recognized, such as three consecutive frames of images.

[0112] Among them, the first historical position recognition result may be the position recognition result of the first recognized image, which is used to indicate whether the object to be recognized is located within the target position.

[0113] Among them, the second recognized image may be a series of images taken within a second preset time period before the shooting time of the image to be recognized. For example, the second preset time includes the first preset time, and the second recognized image includes the first recognized image. For example, the second recognized image may be ten consecutive frames of images.

[0114] Among them, the second historical position recognition result may be the position recognition result of the second recognized image, which is used to indicate whether the object to be recognized is located within the target position.

[0115] Among them, the object state recognition result may be the result of determining whether there is an object to be recognized within the target position based on the position recognition results at multiple time points.

[0116] Optionally, the terminal obtains the first historical position recognition result corresponding to the first recognized image from historical data; determines whether both the first historical position recognition result and the position recognition result are consistent (for example, whether both indicate that the object to be recognized is within the target position); when both the first historical position recognition result and the position recognition result are consistent, obtains the second historical position recognition result corresponding to the second recognized image from historical data; determines whether the position recognition result, the first historical position recognition result, and the second historical position recognition result are all consistent (for example, whether all indicate that the object to be recognized is within the target position), and obtains a comparison result of whether they are all consistent; determines the object state recognition result corresponding to the target position according to this comparison result. For example, when the comparison result is that the position recognition result, the first historical position recognition result, and the second historical position recognition result all indicate that the object to be recognized is within the target position, it is confirmed that the object state recognition result indicates that there is an object to be recognized within the target position, and it can be confirmed that the object to be recognized has successfully located within the target position.

[0117] The technical solution provided in this embodiment compares the historical position recognition results of images at multiple time points with the current position recognition result, so as to more reliably determine whether there is an object to be recognized within the target position, which is beneficial to accurately determining whether there is an object to be recognized within the target position.

[0118] In an exemplary embodiment, after determining the object state recognition result corresponding to the target position, the following is further included: when the object state recognition result indicates that there is an object to be recognized within the target position, generating a transportation instruction corresponding to the object to be recognized; sending the transportation instruction to the transportation device; the transportation device is used to transport the object to be recognized from the target position to the transportation position of the object to be recognized.

[0119] Among them, the transportation instruction may be an instruction for controlling the transportation device to transport the object to be recognized.

[0120] Among them, the transportation device may be a device responsible for transporting the object to be recognized. For example, the transportation device may be an AGV (Automated Guided Vehicle).

[0121] Among them, the transportation position may be the transportation destination position of the object to be recognized.

[0122] Optionally, the terminal determines whether there is an object to be recognized in the target location based on the current location recognition result and the historical location recognition result, and obtains an object status recognition result; if the object status recognition result indicates that there is an object to be recognized in the target location, a corresponding transportation instruction is generated for the object to be recognized, where the instruction content may include information such as the position coordinates of the object to be recognized and the transportation destination (transportation location); the generated transportation instruction is sent to a transportation device (such as an AGV), so that after receiving the transportation instruction, the transportation device autonomously navigates to the target location according to the instruction content and transports the object to be recognized from the target location to the transportation destination.

[0123] The technical solution provided in this embodiment controls the transportation of the object to be recognized through image recognition and judgment, realizes the process of automatically transporting the object to be recognized from one location to another location, and thus is beneficial to improving the transportation efficiency.

[0124] The following uses an embodiment to illustrate the location recognition method provided in this application. This embodiment takes the application of this method to a terminal as an example for illustration. The main steps include:

[0125] First step, the terminal acquires an image to be recognized; the image to be recognized includes an object to be recognized; the position of the object to be recognized in the image to be recognized is recognized through an object position recognition model to obtain the current position of the object to be recognized.

[0126] Second step, the terminal recognizes a candidate area from the image to be recognized; the position of the candidate area is determined as the candidate position of the object to be recognized.

[0127] Third step, when there is an overlap between the current position and the candidate position of the object to be recognized, the terminal extracts a regional image from the image to be recognized; the candidate position that overlaps with the current position is used as the target position.

[0128] Fourth step, the terminal recognizes the vertex position of the object to be recognized in the regional image through a vertex position recognition model to obtain the vertex position of the object to be recognized.

[0129] Fifth step, the terminal recognizes the vertex position and the target position to obtain a vertex position recognition result of the object to be recognized; the vertex position recognition result is used to indicate whether the vertex of the object to be recognized is located within the target position; according to the vertex position recognition result, the current number of vertices located within the target position is determined; when the current number is greater than or equal to a preset number, it is confirmed that the position recognition result of the object to be recognized indicates that the object to be recognized is located within the target position.

[0130] Step 6: The terminal obtains the first historical position recognition result corresponding to the first recognized image; the first recognized image is an image captured within a first preset time period before the shooting time of the image to be recognized; when both the first historical position recognition result and the position recognition result are consistent, the terminal obtains the second historical position recognition result corresponding to the second recognized image; the second recognized image is an image captured within a second preset time period before the shooting time of the image to be recognized; according to the position recognition result, the first historical position recognition result, and the second historical position recognition result, the terminal determines the object state recognition result corresponding to the target position; the object state recognition result is used to indicate whether there is an object to be recognized within the target position.

[0131] Step 7: When the object state recognition result indicates that there is an object to be recognized within the target position, the terminal generates a transportation instruction corresponding to the object to be recognized; the terminal sends the transportation instruction to the transportation device; the transportation device is used to transport the object to be recognized from the target position to the transportation position of the object to be recognized.

[0132] The technical solution provided in this embodiment identifies the current position of the object to be recognized from the image to be recognized. When there is an overlap between the current position and the target position of the object to be recognized, the terminal extracts the region image corresponding to the object to be recognized, and uses the vertex position recognition model to identify the vertex position of the object to be recognized in the region image. By recognizing the vertex position and the target position, the terminal obtains the position recognition result indicating whether the object to be recognized is within the target position. In this way, the solution improves the efficiency and accuracy of position recognition by recognizing the preliminary position and the precise position of the object to be recognized and comparing them with the target position.

[0133] The following uses an application example to illustrate the position recognition method provided in this application. This application example takes the application of this method to a terminal as an example, and the main steps include:

[0134] Step 1: The terminal performs data acquisition: By installing a monitoring camera above the storage location, the terminal monitors multiple storage locations in real time and obtains the real-time image stream data of the site.

[0135] Step 2: The terminal resizes the image to 640*640 and then normalizes it; the terminal sends the image to the RTMDet network (real-time target detection network) and outputs the detected cage car and its coordinate information.

[0136] Among them, the position coordinates of the storage location can be preset in the image in advance.

[0137] Step 3: The terminal obtains the coordinates of the storage location monitored by the monitoring camera, calculates the IOU (intersection over union) between the coordinates of the storage location and the coordinates of the cage car detected in the second step, and extracts the cage car with an IOU>0 with any storage location from the original image.

[0138] For example, the terminal identifies the position of the cage car in the image through the RTMDet network. After identifying that the position of the cage car overlaps with the position of the storage location, the corresponding cage car image is cropped from the original image to obtain the target cage car image.

[0139] Step 4: The terminal resizes the cropped cage car image to 512*512, then normalizes it, and then sends it into the RTMPose (real-time key point detection model) network to detect the four vertices on the top of the cage car, and outputs the coordinate information of the four vertices.

[0140] Step 5: The terminal compares and calculates the coordinate information of the four vertices with the storage location captured by the camera respectively. If at least three of the four vertices are within the coordinate range of the storage location, it means that the cage car is in the storage location (it means that the target cage car is in the target storage location).

[0141] Step 6: The in-place detection of the storage location performed by the terminal cannot rely only on the single-frame result. The error of the single-frame result will cause the instability of the system. Therefore, a state machine is provided for the system of the cage car arriving at the storage location and calling the AGV; specifically: 1. First, set a buffer area for storing the in-place state of the cage car, and the buffer length is 10-frame detection results; 2. When there is a situation where the detection results of three consecutive frames are the same, trace back the results of the previous 10 frames to make the current state determination; the determination logic is that if the results of the three frames are the same as the results of the current state, no change is made. If the results of the three-frame detection are inconsistent with the results of the current state, then when more than 80% of the results in the 10 frames are the same as the results of the last three frames, the current state needs to be changed. If this change is from no cage car to having a cage car, the AGV will be automatically called to transport the cage car.

[0142] For example, the terminal identifies the time when the target cage car is in the target storage location through multiple frames of images. If the time is greater than the preset time threshold, it means that the target cage car has successfully entered the storage; subsequently, through the system of calling the AGV, the transportation of the target cage car is carried out.

[0143] The technical solution provided by this application example realizes improving the efficiency and accuracy of position recognition and realizes automatic cage car calling.

[0144] It should be understood that although the steps in the flowcharts involved in the above-described embodiments are displayed sequentially as indicated by the arrows, these steps are not necessarily executed sequentially in the order indicated by the arrows. Unless there is a clear indication in this article, there is no strict order restriction for the execution of these steps, and these steps can be executed in other orders. Moreover, at least a part of the steps in the flowcharts involved in the above-described embodiments may include multiple steps or multiple stages. These steps or stages are not necessarily executed at the same time, but can be executed at different times. The execution order of these steps or stages is not necessarily sequential, but can be executed alternately or in turn with at least a part of other steps or steps or stages in other steps.

[0145] Based on the same inventive concept, an embodiment of the present application also provides a position recognition device for implementing the above-mentioned position recognition method. The solution provided by this device to solve the problem is similar to the solution described in the above method. Therefore, the specific limitations in one or more embodiments of the position recognition device provided below can refer to the limitations on the position recognition method in the above text, and will not be repeated here.

[0146] In an exemplary embodiment, as Figure 6 shown, a position recognition device is provided. The device 600 may include:

[0147] An object recognition module 601, configured to recognize the current position of the object to be recognized from the acquired image to be recognized;

[0148] An image cropping module 602, configured to crop the region image corresponding to the object to be recognized from the image to be recognized when there is an overlap between the current position and the target position of the object to be recognized;

[0149] A vertex recognition module 603, configured to perform vertex position recognition on the object to be recognized in the region image through a vertex position recognition model to obtain the vertex position of the object to be recognized;

[0150] A target recognition module 604, configured to perform recognition on the vertex position and the target position to obtain a position recognition result of the object to be recognized; the position recognition result is used to indicate whether the object to be recognized is located within the target position.

[0151] In an exemplary embodiment, the target recognition module 604 is further configured to recognize the vertex position and the target position to obtain a vertex position recognition result of the object to be recognized; the vertex position recognition result is used to indicate whether the vertex of the object to be recognized is within the target position; according to the vertex position recognition result, determine the current number of vertices located within the target position; in the case where the current number is greater than or equal to a preset number, confirm that the position recognition result of the object to be recognized indicates that the object to be recognized is located within the target position.

[0152] In an exemplary embodiment, the object recognition module 601 is further configured to obtain an image to be recognized; the image to be recognized includes an object to be recognized; perform position recognition on the object to be recognized in the image to be recognized through an object position recognition model to obtain the current position of the object to be recognized.

[0153] In an exemplary embodiment, the image cropping module 602 is further configured to crop a region image from the image to be recognized in the case where there is an overlap between the current position and the candidate position of the object to be recognized; use the candidate position that overlaps with the current position as the target position.

[0154] In an exemplary embodiment, the apparatus 600 further includes: a position determination module, configured to recognize a candidate region from the image to be recognized; determine the position of the candidate region as the candidate position of the object to be recognized.

[0155] In an exemplary embodiment, the apparatus 600 further includes: a result determination module, configured to obtain a first historical position recognition result corresponding to a first recognized image; the first recognized image is an image captured within a first preset time period before the shooting time of the image to be recognized; in the case where the first historical position recognition result and the position recognition result are both consistent, obtain a second historical position recognition result corresponding to a second recognized image; the second recognized image is an image captured within a second preset time period before the shooting time of the image to be recognized; determine an object state recognition result corresponding to the target position according to the position recognition result, the first historical position recognition result, and the second historical position recognition result; the object state recognition result is used to indicate whether there is an object to be recognized within the target position.

[0156] In an exemplary embodiment, the apparatus 600 further includes: an instruction sending module, configured to generate a transportation instruction corresponding to the object to be recognized in the case where the object state recognition result indicates that there is an object to be recognized within the target position; send the transportation instruction to a transportation device; the transportation device is configured to transport the object to be recognized from the target position to the transportation position of the object to be recognized.

[0157] Each module in the above position recognition device can be implemented in whole or in part by software, hardware, or a combination thereof. Each of the above modules can be embedded in the processor of the computer device in hardware form or be independent of it, or can be stored in the memory of the computer device in software form so that the processor can call and execute the operations corresponding to each of the above modules.

[0158] In an exemplary embodiment, a computer device is provided. The computer device can be a terminal, and its internal structure diagram can be as Figure 7 shown. The computer device includes a processor, a memory, an input / output interface, a communication interface, a display unit, and an input device. Among them, the processor, the memory, and the input / output interface are connected through a system bus, and the communication interface, the display unit, and the input device are connected to the system bus through the input / output interface. Among them, the processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The input / output interface of the computer device is used for the processor to exchange information with external devices. The communication interface of the computer device is used to communicate with external terminals in a wired or wireless manner, and the wireless manner can be achieved through WIFI, a mobile cellular network, NFC (Near Field Communication), or other technologies. The computer program, when executed by the processor, implements a position recognition method. The display unit of the computer device is used to form a visually visible picture, which can be a display screen, a projection device, or a virtual reality imaging device. The display screen can be a liquid crystal display screen or an electronic ink display screen. The input device of the computer device can be a touch layer covering the display screen, or a button, a trackball, or a touchpad provided on the shell of the computer device, or an external keyboard, a touchpad, or a mouse, etc.

[0159] Those skilled in the art can understand that Figure 7 the structure shown in

[0160] is only a block diagram of some structures related to the solution of this application, and does not constitute a limitation on the computer device to which the solution of this application is applied. The specific computer device may include more or fewer components than those shown in the figure, or combine some components, or have different component arrangements.

[0161] In an exemplary embodiment, a computer-readable storage medium is provided, on which a computer program is stored. When the computer program is executed by a processor, the steps in the above method embodiments are implemented.

[0162] In an exemplary embodiment, a computer program product is provided, including a computer program. When the computer program is executed by a processor, the steps in the above method embodiments are implemented.

[0163] Those of ordinary skill in the art can understand that all or part of the processes in the above method embodiments can be completed by instructing relevant hardware through a computer program. The computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the above method embodiments. Among them, any reference to a memory, database, or other medium used in the embodiments provided in the present application can include at least one of non-volatile and volatile memories. Non-volatile memories can include read-only memory (ROM), magnetic tapes, floppy disks, flash memories, optical memories, high-density embedded non-volatile memories, resistive random access memories (ReRAM), magnetoresistive random access memories (MRAM), ferroelectric random access memories (FRAM), phase change memories (PCM), graphene memories, etc. Volatile memories can include random access memory (RAM) or external cache memories, etc. By way of illustration and not limitation, RAM can be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM), etc. The databases involved in the embodiments provided in the present application can include at least one of relational databases and non-relational databases. Non-relational databases can include distributed databases based on blockchain, etc., without limitation. The processors involved in the embodiments provided in the present application can be general-purpose processors, central processors, graphics processors, digital signal processors, programmable logic devices, data processing logics based on quantum computing, etc., without limitation.

[0164] The technical features of the above embodiments can be combined arbitrarily. For the sake of brevity of description, 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, it should be considered to be within the scope described in this specification.

[0165] The embodiments described above merely represent several implementation manners of the present application. The description thereof is relatively specific and detailed, but it should not be construed as a limitation to the patent scope of the present application. It should be noted that for those of ordinary skill in the art, without departing from the concept of the present application, several modifications and improvements can still be made, 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 appended claims.

Claims

1. A position recognition method, characterized in that, The method includes: Identifying the current position of the object to be recognized from the acquired image to be recognized; When there is an overlap between the current position and the target position of the object to be recognized, cropping out the regional image corresponding to the object to be recognized from the image to be recognized; Identifying the vertex positions of the object to be recognized in the regional image through a vertex position recognition model to obtain the vertex positions of the object to be recognized; Identifying the vertex positions and the target position to obtain a position recognition result of the object to be recognized; the position recognition result is used to indicate whether the object to be recognized is within the target position.

2. The method according to claim 1, wherein The identifying the vertex positions and the target position to obtain a position recognition result of the object to be recognized includes: Identifying the vertex positions and the target position to obtain a vertex position recognition result of the object to be recognized; the vertex position recognition result is used to indicate whether the vertices of the object to be recognized are within the target position; Determining the current number of the vertices located within the target position according to the vertex position recognition result; When the current number is greater than or equal to a preset number, confirming that the position recognition result of the object to be recognized indicates that the object to be recognized is within the target position.

3. The method according to claim 1, wherein The identifying the current position of the object to be recognized from the acquired image to be recognized includes: Acquiring the image to be recognized; the image to be recognized includes the object to be recognized; Identifying the position of the object to be recognized in the image to be recognized through an object position recognition model to obtain the current position of the object to be recognized.

4. The method according to claim 1, wherein The cropping out the regional image corresponding to the object to be recognized from the image to be recognized when there is an overlap between the current position and the target position of the object to be recognized includes: When there is an overlap between the current position and a candidate position of the object to be recognized, cropping out the regional image from the image to be recognized; Taking the candidate position that overlaps with the current position as the target position.

5. The method according to claim 4, wherein The method further includes: Identifying a candidate region from the image to be recognized; Determining the position of the candidate region as the candidate position of the object to be recognized.

6. The method according to claim 1, characterized in that After identifying the vertex positions and the target position to obtain a position recognition result of the object to be recognized, the method further includes: Acquiring a first historical position recognition result corresponding to a first recognized image; the first recognized image is an image taken within a first preset time period before the shooting time of the image to be recognized; When the first historical position recognition result and the position recognition result are both consistent, acquiring a second historical position recognition result corresponding to a second recognized image; the second recognized image is an image taken within a second preset time period before the shooting time of the image to be recognized; Based on the position recognition result, the first historical position recognition result, and the second historical position recognition result, determine the object state recognition result corresponding to the target position; the object state recognition result is used to indicate whether the object to be recognized is in the target position.

7. The method according to claim 6, characterized in that, After determining the object state recognition result corresponding to the target position, the method further includes: When the object state recognition result indicates that the object to be recognized is in the target position, generate a transportation instruction corresponding to the object to be recognized; Send the transportation instruction to a transportation device; the transportation device is used to transport the object to be recognized from the target position to the transportation position of the object to be recognized.

8. A position recognition device, characterized in that, The device includes: An object recognition module, configured to recognize the current position of the object to be recognized from the acquired image to be recognized; An image cropping module, configured to crop out the region image corresponding to the object to be recognized from the image to be recognized when there is an overlap between the current position and the target position of the object to be recognized; A vertex recognition module, configured to perform vertex position recognition on the object to be recognized in the region image through a vertex position recognition model to obtain the vertex position of the object to be recognized; A target recognition module, configured to recognize the vertex position and the target position to obtain a position recognition result of the object to be recognized; the position recognition result is used to indicate whether the object to be recognized is in the target position.

9. A computer device, comprising a memory and a processor, the memory storing a computer program, characterized in that, When the processor executes the computer program, the steps of the method according to any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 7 are implemented.

11. A computer program product, comprising a computer program, characterized in that, When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 7 are implemented.

Images