Robot control method and device, storage medium and electronic equipment

Abstract

The invention relates to a robot control method and device, a storage medium and electronic equipment, in particular to the technical field of intelligent warehousing, and the method comprises the steps that firstly, attribute information of a carried object is acquired; analyzing the attribute information, and determining a container model corresponding to the carrying object according to an analysis result; then obtaining a model identifier of the container model; and a carrying instruction of the carried object is sent to the robot. According to the technical scheme, the carrying task is issued to the robot according to the model identifier of the container model corresponding to the carrying object, and the robot can select the suitable camera recognition mode according to the model identifier in the carrying instruction, quickly and effectively recognize the edge of the carrying object and make an accurate action response accordingly, so that the carrying efficiency is improved. The time consumed for adjusting camera parameters and identifying the edge of the carried object is saved, the efficiency of carrying tasks executed by the robot is improved, and the flexibility and robustness of the system are enhanced.

Description

Technical Field

[0001] The present disclosure relates to the technical field of intelligent warehousing, and particularly relates to a robot control method, device, storage medium and electronic device. Background Art

[0002] With the development of intelligent warehousing technology, robots have been widely used. Robots can carry handling objects such as cargo boxes and shelves.

[0003] Currently, after sending a handling instruction for a handling object to a robot, if the edge of the handling object is not easily recognizable, such as being interfered by factors such as the color and reflectivity of the surface of the handling object, the robot needs to adjust the camera parameters multiple times. After identifying the edge of the handling object, it can execute the corresponding handling task. This method takes a relatively long time and thus affects the efficiency of the robot in executing the handling task. Summary of the Invention

[0004] In view of this, the present disclosure provides a robot control method, device, storage medium and electronic device, mainly aiming to improve the technical problem that in the current related technologies, the method of the robot needing to adjust the camera parameters multiple times and then executing the corresponding handling task after identifying the edge of the handling object takes a relatively long time, thereby affecting the efficiency of the robot in executing the handling task.

[0005] In a first aspect, the present disclosure provides a robot control method, including:

[0006] Obtain the attribute information of the handling object;

[0007] Analyze the attribute information, and determine the container model corresponding to the handling object according to the analysis result;

[0008] Obtain the model identifier of the container model;

[0009] Send a handling instruction for the handling object to the robot, where the handling instruction carries the model identifier, and the model identifier is used for the robot to determine the camera recognition mode used when recognizing the handling object.

[0010] Optionally, the determining the container model corresponding to the handling object according to the analysis result includes:

[0011] Obtain the object identifier of the handling object from the analysis result;

[0012] Determine the model identifier according to the object identifier.

[0013] Optionally, the determining the model identifier according to the object identifier includes:

[0014] Extract the target character at the preset position of the object identifier;

[0015] Determine the model identifier corresponding to the target character according to the preset mapping rule between the character and the model identifier.

[0016] Optionally, before obtaining the model identifier of the container model corresponding to the handling object, the method further includes:

[0017] Obtain the appearance features of different handling objects;

[0018] Configure the container models respectively corresponding to different handling objects according to the appearance features;

[0019] Determine the camera recognition modes respectively corresponding to different container models, and determine the preset mapping rule between the model identifiers of different container models and different characters.

[0020] Optionally, determining the camera recognition mode corresponding to the container model includes:

[0021] Obtain the model parameters corresponding to the container model, where the model parameters include one or more of the size, color, material, and surface texture of the container model;

[0022] Determine the camera parameters based on the model parameters and the preset working environment, where the camera parameters include one or more of the focal length, aperture size, sensitivity, exposure time, and white balance;

[0023] Perform a test shooting on the model sample of the container model according to the camera parameters;

[0024] Determine the camera recognition mode corresponding to the container model according to the test shooting result.

[0025] Optionally, the determining the camera recognition mode corresponding to the container model according to the test shooting result includes:

[0026] Perform image quality analysis on the captured images of the model sample under different camera parameter conditions respectively;

[0027] Obtain the target camera parameters corresponding to the target captured image with the image quality greater than the preset threshold;

[0028] Determine the camera recognition mode corresponding to the container model according to the target camera parameters.

[0029] Optionally, the determining the camera recognition mode corresponding to the container model according to the target camera parameters includes:

[0030] Preprocess the target captured image;

[0031] Perform edge recognition on the preprocessed target captured image through different vision algorithms;

[0032] Obtain the target vision algorithm whose edge recognition result meets the requirements;

[0033] According to the target vision algorithm and the target camera parameters, obtain the camera recognition mode corresponding to the container model.

[0034] Optionally, the method further includes:

[0035] Send the mapping relationship between the model identifiers of different container models and the camera recognition modes to the robot.

[0036] Optionally, after sending the handling instruction of the handling object to the robot, the method further includes:

[0037] Receive the target camera recognition mode used when the edge recognition of the handling object by the robot is successful;

[0038] Update the mapping relationship between the model identifier of the container model and the camera recognition mode according to the target camera recognition mode;

[0039] Send the updated mapping relationship to the robot.

[0040] Optionally, before determining the container model corresponding to the handling object according to the parsing result, the method further includes:

[0041] Perform interface verification according to the parsing result to determine whether there is a container model corresponding to the handling object;

[0042] If there is no container model corresponding to the handling object, it is determined that adding the handling object to the database fails and an appropriate error message is triggered for output, where the objects successfully added to the database can be tracked, scheduled, and managed;

[0043] The determining the container model corresponding to the handling object according to the parsing result includes:

[0044] If there is a container model corresponding to the handling object, add the handling object to the database and determine the container model corresponding to the handling object.

[0045] Optionally, performing interface verification according to the parsing result to determine whether there is a container model corresponding to the handling object includes:

[0046] Obtain the object identifier of the handling object from the parsing result;

[0047] Determine whether the object identifier conforms to a preset identifier form, where the identifier of the preset identifier form has a corresponding container model;

[0048] If the object identifier conforms to the preset identifier form, it is determined that there is a container model corresponding to the handling object;

[0049] If the object identifier does not conform to the preset identifier form, it is determined that there is no container model corresponding to the handling object.

[0050] In a second aspect, the present disclosure provides a robot control method, including:

[0051] Receive a handling instruction for a handling object, where the handling instruction carries a model identifier of a container model corresponding to the handling object, and the container model is determined according to the parsing result of the attribute information of the handling object;

[0052] According to the model identifier, determine the camera recognition mode used by the robot when recognizing the handling object.

[0053] Optionally, before determining the camera recognition mode used by the robot when recognizing the handling object according to the model identifier, the method further includes:

[0054] Receive the mapping relationship between the model identifiers of different container models and the camera recognition modes;

[0055] According to the model identifier, determining the camera recognition mode used by the robot when recognizing the handling object includes:

[0056] Determine the camera recognition mode used by the robot when recognizing the handling object by querying the mapping relationship.

[0057] Optionally, according to the model identifier, determining the camera recognition mode used by the robot when recognizing the handling object includes:

[0058] Determine the camera parameters and vision algorithms used by the robot when recognizing the handling object.

[0059] Optionally, the method further includes:

[0060] Perform edge recognition on the handling object based on the camera parameters and vision algorithms.

[0061] Optionally, performing edge recognition on the handling object based on the camera parameters and vision algorithms includes:

[0062] Use the robot camera device to capture the handling object according to the camera parameters;

[0063] Perform edge recognition on the captured image according to the visual algorithm to obtain an edge map marked with the edge positions.

[0064] Match the edge features in the edge map with the edge features of the container model corresponding to the model identifier.

[0065] Determine whether the edge recognition is successful based on the matching result, and obtain the successfully recognized edge positions in the case of successful edge recognition.

[0066] Optionally, the method further includes:

[0067] During the process of recognizing the handling object using the camera recognition mode determined according to the model identifier, if the edge recognition fails after a preset number of attempts, then use the replaced camera recognition mode to recognize the handling object.

[0068] Send the target camera recognition mode used when the edge recognition of the handling object is successful to update the mapping relationship between the model identifier of the container model and the camera recognition mode.

[0069] In a third aspect, the present disclosure provides a robot control device, including:

[0070] An acquisition module, configured to acquire the attribute information of the handling object; parse the attribute information, and determine the container model corresponding to the handling object according to the parsing result; acquire the model identifier of the container model.

[0071] A sending module, configured to send a handling instruction for the handling object to the robot, where the handling instruction carries the model identifier, and the model identifier is used for the robot to determine the camera recognition mode used when recognizing the handling object.

[0072] In a fourth aspect, the present disclosure provides a robot control device, including:

[0073] A receiving module, configured to receive a handling instruction for the handling object, where the handling instruction carries the model identifier of the container model corresponding to the handling object, and the container model is determined according to the parsing result of the attribute information of the handling object.

[0074] A determining module, configured to determine the camera recognition mode used by the robot when recognizing the handling object according to the model identifier.

[0075] In a fifth aspect, the present disclosure provides a computer-readable storage medium, on which a computer program is stored, and when the computer program is executed by a processor, it implements the robot control method described in the first aspect or the second aspect.

[0076] In a sixth aspect, the present disclosure provides an electronic device, including a storage medium, a processor, and a computer program stored on the storage medium and executable on the processor. When the processor executes the computer program, the robot control method described in the first aspect or the second aspect is implemented.

[0077] In a seventh aspect, the present disclosure provides a computer program product, including a computer program which, when executed by a processor, implements the robot control method described in the first aspect or the second aspect.

[0078] By means of the above technical solutions, a robot control method, device, storage medium, and electronic device provided by the present disclosure first obtain attribute information of a handling object; parse the attribute information and determine a container model corresponding to the handling object according to the parsing result; then obtain a model identifier of the container model; and then send a handling instruction for the handling object to the robot, where the handling instruction carries the model identifier, and the model identifier is used for the robot to determine the camera recognition mode used when recognizing the handling object. Compared with the current related technologies, according to the model identifier of the container model corresponding to the handling object, the present disclosure issues a handling task to the robot. The robot can select an applicable camera recognition mode according to the model identifier in the handling instruction, quickly and effectively identify the edge of the handling object, and make an accurate action response accordingly, saving the time consumed for adjusting camera parameters and identifying the edge of the handling object, improving the efficiency of the robot in executing the handling task, and enhancing the flexibility and robustness of the system.

[0079] The above description is only an overview of the technical solutions of the present disclosure. In order to be able to understand the technical means of the present disclosure more clearly, it can be implemented according to the content of the description. And in order to make the above and other purposes, features, and advantages of the present disclosure more obvious and understandable, the specific embodiments of the present disclosure are specifically exemplified below. BRIEF DESCRIPTION OF THE DRAWINGS

[0080] The drawings herein are incorporated into the description and constitute a part of this description, showing embodiments consistent with the present disclosure and used together with the description to explain the principles of the present disclosure.

[0081] In order to more clearly illustrate the technical solutions in the embodiments of the present disclosure or the prior art, the following will briefly introduce the drawings required for use in the description of the embodiments or the prior art. Obviously, for those of ordinary skill in the art, other drawings can be obtained according to these drawings without creative efforts.

[0082] Figure 1 A flowchart showing a robot control method provided by an embodiment of the present disclosure is shown;

[0083] Figure 2Shows a schematic diagram of an example provided by an embodiment of the present disclosure;

[0084] Figure 3 Shows a schematic flowchart of a robot control method provided by an embodiment of the present disclosure;

[0085] Figure 4 Shows a schematic diagram of an example provided by an embodiment of the present disclosure;

[0086] Figure 5 Shows a schematic flowchart of a robot control method provided by an embodiment of the present disclosure;

[0087] Figure 6 Shows a schematic structural diagram of a robot control device provided by an embodiment of the present disclosure;

[0088] Figure 7 Shows a schematic structural diagram of a robot control device provided by an embodiment of the present disclosure;

[0089] Figure 8 Shows a schematic structural diagram of an electronic device provided by an embodiment of the present disclosure. Detailed implementation manners

[0090] In order to more clearly understand the above objects, features and advantages of the present disclosure, the solutions of the present disclosure will be further described below. It should be noted that, without conflict, the embodiments of the present disclosure and the features in the embodiments may be combined with each other.

[0091] Many specific details are set forth in the following description in order to provide a thorough understanding of the present disclosure. However, the present disclosure can be implemented in many other ways different from those described herein, and those skilled in the art can make similar extensions without departing from the connotation of the present disclosure. Therefore, the present disclosure is not limited by the specific implementations disclosed below.

[0092] The terms used in one or more embodiments of the present disclosure are only for the purpose of describing specific embodiments, and are not intended to limit one or more embodiments of the present disclosure. The singular forms "a", "the" and "said" used in one or more embodiments of the present disclosure and the appended claims are also intended to include the plural forms, unless the context clearly indicates otherwise. It should also be understood that the term " / and / " used in one or more embodiments of the present disclosure refers to and includes any or all possible combinations of one or more related listed items.

[0093] It should be understood that although the terms first, second, etc. may be used in one or more embodiments of the present disclosure to describe various information, such information should not be limited to these terms. These terms are only used to distinguish information of the same type from each other. For example, without departing from the scope of one or more embodiments of the present disclosure, the first may also be referred to as the second, and similarly, the second may also be referred to as the first. Depending on the context, the word "if" as used herein may be interpreted as "when" or "while" or "in response to determining".

[0094] To improve the technical problem in the current related art that the robot needs to adjust the camera parameters multiple times, and the method of performing the corresponding handling task after identifying the edge of the handling object takes a relatively long time, thereby affecting the efficiency of the robot in performing the handling task. An embodiment of the present disclosure provides a robot control method, as Figure 1 shown, the method includes the following steps 101 to step 104:

[0095] Step 101, obtain the attribute information of the handling object.

[0096] In some examples, the handling object may be an object such as a vehicle, a container, a document, a cargo, etc. Among them, when the handling object is a vehicle, the handling object may also be a shelf. When the handling object is a container, the handling object may be a pallet, a box body, a document bag, a flexible intermediate bulk container or a packaging bag, etc. Among them, the box body may be a wooden box, a cardboard box or a plastic box, etc. The attribute information is used to determine the attribute of the handling object, and may include one or more of the three-dimensional size, weight, shape description, center of gravity position, surface characteristics, coding, etc. of the handling object.

[0097] Step 102, parse the attribute information, and determine the container model corresponding to the handling object according to the parsing result.

[0098] In some examples, the container model can accurately display the shape and size of the handling object, including its height, diameter, the shape of the bottom and the top, etc., which helps the user to understand the physical size of the handling object for space planning or logistics transportation, etc.

[0099] Exemplarily, by parsing the attribute information, according to the size, weight and shape of the handling object, determine the container type that can accommodate and protect the handling object.

[0100] Optionally, step 102 may specifically include: obtaining the object identifier of the handling object from the parsing result; determining the model identifier according to the object identifier.

[0101] Exemplarily, the Warehouse Management System (WMS) determines the handling objects to be moved according to order requirements or inventory management strategies, generates corresponding handling tasks, and sends handling instructions to the Robot Management System (RMS). The task instructions carry object identifiers (such as bin numbers, barcodes, QR codes, labels, etc.). The RMS further resolves the model identifier by querying the internal database or other means based on the object identifier. Among them, the object identifier can be used to identify each handling object in the warehouse. The model identifier corresponds to the container model, and the model identifier can be used to indicate the color, specifications, dimensions, etc. of the handling object.

[0102] In some examples, to simplify the process or improve the response speed, the task instructions issued by the WMS can also directly carry the model identifier. In this way, the RMS does not need to perform additional parsing work and can directly assign appropriate robots to execute the task, improving work efficiency and enabling the robot to better adapt to different types of handling objects.

[0103] Optionally, the determination of the model identifier according to the object identifier may specifically include: extracting target characters at preset positions of the object identifier; and determining the model identifier corresponding to the target characters according to the preset mapping rule between the characters and the model identifier.

[0104] Exemplarily, one or more preset positions can be defined to extract specific target characters in the object identifier. These characters can be numbers, letters, or a combination of them. For example, if the object identifier is a string, these target characters can be found at fixed positions (such as the first two digits, the last three digits, etc.) of the string. A set of preset mapping rules between the characters and the model identifier can be established. The RMS resolves the corresponding model identifier based on the target characters in the object identifier of the handling object according to the preset mapping rule, and then determines the model identifier.

[0105] Exemplarily, one bin number cannot correspond to multiple model identifiers, and multiple bin numbers can correspond to one model identifier, ensuring that each bin number corresponds to only one unique model identifier and avoiding confusion and errors. For the case where multiple bin numbers correspond to one model identifier, it means that there is a group of bin numbers sharing the same physical or functional characteristics, such as size, weight limit, etc. Therefore, when the RMS receives the task instructions, it only needs to identify the key part (i.e., the target characters) in the bin number to quickly and accurately determine which identification or processing method should be used to handle these bins.

[0106] For example, the target characters at the preset position of the object identifier can be defined as the first two digits of the container number. For example, the first two digits of the container number of the black container are 71, that is, it is stipulated that the containers with container numbers from 71000000 to 71999999 are black containers, and the container model is the model corresponding to the black container. In this configuration method, in order to obtain the model identifier of the container, the first two digits of the container number can be parsed. In this process, first, a clear mapping relationship between the container number and the container model needs to be established. Specifically, a mapping table or rule set can be created to associate specific target characters (i.e., the first two digits of the container number) with the model identifier. For example, the container with the first two digits of the container number being 71 is mapped to the model identifier "1", which means that these containers correspond to the black container model. When the first two digits of the container number are 71, its model identifier is "1". Therefore, if the first two digits of the container number in the task assigned by the WMS are 71, it is equivalent to the model identifier of the container used in this task being 1. Through the preset mapping rules between characters and model identifiers, a large number of handling objects and their corresponding models can be effectively managed, and the speed and accuracy of task processing are also improved.

[0107] Optionally, the method of this embodiment may further specifically include: performing interface verification according to the parsing result to determine whether there is a container model corresponding to the handling object; if there is no container model corresponding to the handling object, it is determined that adding the handling object to the database fails and an appropriate error message is triggered for output, where the objects successfully added to the database can be tracked, scheduled, and managed; correspondingly, step 102 may specifically include: if there is a container model corresponding to the handling object, adding the handling object to the database and determining the container model corresponding to the handling object.

[0108] In some embodiments, when entering a container through the addbox interface, the system stores the model identifier matching the container model in the database of the RMS, ensuring that the RMS can accurately parse the corresponding container model according to the container number in subsequent operations.

[0109] In some examples, if the parameters related to the model identifier are not configured in the RMS, the parameter parsing step is skipped, and the system will continue with the original operation process; if the relevant parameters are configured in the RMS, based on the parsed parameters, the RMS will retrieve the corresponding model identifier according to the container number. For example, the first two digits of the container number are extracted, and the model identifier is determined according to the preset mapping rule. Once the matching model identifier is found, the system will assign it to the preset target field. Similarly, during the interface verification process, the RMS will take different measures according to the configuration: if the relevant parameters are not configured, the RMS will skip the interface verification step; if the parameters are configured and the container number meets the conditions set by the parameters, the RMS will perform additional verification: check whether the container model corresponding to the model identifier exists in the database. If the model does not exist, that is, the RMS cannot find the model identifier corresponding to the container number in the database, the system will report an error and return an error code, indicating that "the container model number does not exist". In this case, the operation of adding a container will fail; or it is also possible to detect whether there is an error in the model identifier of the container model in the RMS and / or the container number sent by the WMS, and modify the model identifier or the container number sent by the WMS according to the detection result.

[0110] Optionally, the above interface verification based on the parsing result to determine whether there is a container model corresponding to the handling object may specifically include: obtaining the object identifier of the handling object from the parsing result; determining whether the object identifier conforms to a preset identifier form, where the identifier of the preset identifier form has a corresponding container model; if the object identifier conforms to the preset identifier form, it is determined that there is a container model corresponding to the handling object; if the object identifier does not conform to the preset identifier form, it is determined that there is no container model corresponding to the handling object.

[0111] Exemplarily, the preset identifier form can be defined according to business requirements or system specifications, and is used to stipulate requirements such as the format, length, and character type of the object identifier. These requirements are usually related to the corresponding relationship of the container model. The attribute information of the handling object can be parsed, and the object identifier is found and extracted from the parsed attribute information. If the object identifier conforms to the preset identifier form, then it can be considered that the object identifier is valid and has a corresponding relationship with a certain container model. At this time, the corresponding container model information can be found according to the object identifier. If the object identifier does not conform to the preset identifier form, then it can be considered that the object identifier is invalid, or at least there is no corresponding container model in the current system. At this time, further processing may be required, such as prompting the user to enter the correct object identifier, querying other relevant information, or performing error handling, etc.

[0112] Step 103, obtain the model identifier of the container model.

[0113] Exemplarily, a model identifier (model ID) can be used to distinguish different container models. For example, container models can include black cargo box models, red cargo box models, dark gray cargo box models, navy blue cargo box models, etc. Different container models correspond to different model identifiers. For example, the model identifier corresponding to the black cargo box model is 1, the model identifier corresponding to the red cargo box model is 2, the model identifier corresponding to the dark gray cargo box model is 3, the model identifier corresponding to the navy blue cargo box model is 4, and so on.

[0114] Step 104: Send a handling instruction for the handling object to the robot.

[0115] Among them, the handling instruction carries a model identifier, which is used by the robot to determine the camera recognition mode used when identifying the handling object. Different camera recognition modes can be used to identify different container models. Selecting the camera recognition mode according to the determination logic corresponding to the model identifier can quickly perform the identification of the handling object.

[0116] Exemplarily, the handling instruction can include a box-taking instruction, a box-placing instruction, etc. The box-taking instruction is used to instruct the robot to move the handling object out of its current location, and the box-placing instruction is used to instruct the robot to move the handling object to a specified target location and place it safely and accurately.

[0117] In some examples, when RMS sends a box-taking and placing instruction to the robot, it can indicate the model identifier through a certain byte in the instruction. For example, it can use bytes 0-4 in the object identifier to indicate the model identifier and bytes 5-7 to indicate the model color. If the model identifier is "1", it represents that the corresponding container model is the model corresponding to the black cargo box; correspondingly, if the model identifier is "0", it represents that the corresponding container model is the model corresponding to the non-black cargo box.

[0118] For example, as Figure 2 shown, the WMS sends the cargo box number to the RMS. The RMS resolves the corresponding model identifier through the cargo box number and, when sending a box-taking task to the robot, sends the box-taking instruction carrying the model identifier to the robot. The robot can then quickly perform the identification of the handling object according to the determination logic corresponding to the model identifier. For example, select the recognition mode of the depth camera according to the cargo box color corresponding to the model identifier, and perform the box-taking and placing tasks of the black cargo box in the recognition mode of the black cargo box (such as relaxing the motor adjustment to increase the compensation value, etc.); perform the box-taking and placing tasks of the non-black cargo box in the recognition mode of the non-black cargo box, etc.

[0119] In some examples, for red cargo box types, dark gray cargo box types, navy blue cargo box types, etc., there are also their respective corresponding recognition modes.

[0120] Compared with the current related technologies, in the embodiments of the present disclosure, the attribute information of the handling object is first obtained; the attribute information is parsed, and the container model corresponding to the handling object is determined according to the parsing result; then the model identifier of the container model is obtained; and then a handling instruction for the handling object is sent to the robot, where the handling instruction carries the model identifier, and the model identifier is used for the robot to determine the camera recognition mode used when recognizing the handling object. By applying the method of this embodiment, according to the model identifier of the container model corresponding to the handling object, the handling task is sent to the robot. The robot can determine the container model corresponding to the handling object according to the model identifier in the handling instruction, and select a suitable camera recognition mode, quickly and effectively recognize the edge of the handling object, and make an accurate action response accordingly, saving the time consumed for adjusting the camera parameters and recognizing the edge of the handling object, improving the efficiency of the robot in executing the handling task, and enhancing the flexibility and robustness of the system.

[0121] Further, as Figure 1 an optional manner of the embodiment, as Figure 3 shown, before obtaining the model identifier of the container model corresponding to the handling object, it is necessary to establish a preset mapping rule between the model identifier and different characters, which may include the methods shown in steps 201 to 203 as follows:

[0122] Step 201: Obtain the appearance features of different handling objects.

[0123] For example, multi-angle pictures of the handling object can be taken first using a depth camera to ensure that all important features are covered, and key features such as the size, color, shape, and texture of the object are extracted through computer vision algorithms or deep learning models.

[0124] Step 202: Configure the container models corresponding to different handling objects according to the appearance features.

[0125] Exemplarily, corresponding container models can be created according to the extracted appearance features. These models can be 3D models or simplified geometric shapes, such as Figure 4 the cargo box shown in, and by parametrically defining the size, material, and other physical properties of the container model, it can accurately display the shape, size, etc. of the handling object.

[0126] Step 203: Determine the camera recognition modes corresponding to different container models, and determine the preset mapping rule between the model identifiers of different container models and different characters.

[0127] In some examples, different camera recognition modes can be used to recognize different container models. By selecting the camera recognition mode according to the corresponding determination logic of the model identification, the recognition of handling objects can be quickly performed. For example, an appropriate recognition algorithm can be selected based on color and / or texture to distinguish handling objects with obvious color differences or unique surface textures; or an appropriate recognition algorithm can be selected based on shape and / or contour for recognizing objects with regular shapes and easy to identify; or based on depth information, three-dimensional reconstruction can be performed using the depth data provided by a depth camera for the recognition of objects with complex shapes, etc.

[0128] Exemplarily, a preset mapping rule can be used to convert the characters representing the container model into the corresponding model identification. In one example, first, a unique model identification or number needs to be assigned to each container model as its identity identifier, and then a set of character sets that are easy to read and process (such as ASCII code) is selected to represent different model identifications. By creating a mapping rule table, the model identification of each container model is associated with its corresponding character code. For example, container model A corresponds to the character "0", and container model B corresponds to the character "1".

[0129] Optionally, the method of this embodiment may further include: obtaining the model parameters corresponding to the container model, where the model parameters include one or more of the size, color, material, and surface texture of the container model; determining the camera parameters based on the model parameters and the preset working environment, where the camera parameters include one or more of the focal length, aperture size, sensitivity, exposure time, and white balance; performing a test shooting on the model sample of the container model according to the camera parameters; and determining the camera recognition mode corresponding to the container model based on the test shooting result.

[0130] In some examples, camera parameters are determined based on model parameters and a preset working environment. Specifically, an appropriate lens focal length can be selected according to the size of the container model and the shooting distance, the aperture value can be adjusted to control the depth of field, ensuring that the key features of the container model are within the focus range. The appropriate ISO can be set according to the environmental light conditions to balance the image quality and noise level, and the shutter speed can be adjusted to prevent overexposure or underexposure of the image due to too long or too short exposure. By calibrating the white balance setting, accurate color reproduction can be ensured, especially under multi-light source or complex lighting conditions. Based on machine learning or deep learning methods, a sample data set can also be collected to train the model so that it can distinguish different container edges. Photos of the container are taken from multiple angles according to the determined camera parameters to obtain images of the container under different conditions, covering all important features, and detailed annotation is carried out. Image processing software or a deep learning model is used to automatically evaluate the quality of each photo, check for problems such as blurring, distortion, and overexposure, and extract key features from the qualified test images to verify whether different types of container models can be effectively distinguished. Iterative optimization is continuously carried out according to the problems found in the verification process until the model can stably and reliably help the robot identify the container edge. If it is found that certain parameter settings cause image quality problems, corresponding adjustments are required, and the test is repeated until satisfactory results are obtained.

[0131] Exemplarily, after the test shooting, factors such as image quality and feature extraction success rate are comprehensively considered to check whether the image quality meets the requirements, especially the performance in terms of edge sharpness, contrast, etc. The effectiveness of the current camera parameter settings is evaluated, and a set of optimal camera parameter combinations is selected according to the test results to ensure that various container models can be stably identified in practical applications. For a specific type of container model, the most suitable edge detection algorithm is selected according to the characteristics of the container and the test results. Commonly used algorithms include Canny edge detection, Sobel operator, Laplacian operator, depth algorithms, etc.; for more complex scenarios, deep learning methods can also be considered to improve the recognition accuracy and speed. Selecting the most suitable recognition algorithm and parameter combination can ensure that the system can operate stably in the actual environment.

[0132] Optionally, determining the camera recognition mode corresponding to the container model based on the test shooting results may specifically include: performing image quality analysis on the captured images of the model samples under different camera parameters; obtaining the target camera parameters corresponding to the target captured images with image quality greater than a preset threshold; and determining the camera recognition mode corresponding to the container model based on the target camera parameters.

[0133] In some examples, to ensure the recognition accuracy and stability of the container model in the robot handling system, a reasonable image quality threshold can be set according to industry specifications or actual requirements (such as clarity > 70 points, color accuracy < 5ΔE, etc.). If the preliminary evaluation finds that the threshold is too high or too low, it can be appropriately adjusted according to the actual situation. Based on the above evaluation indicators, the target captured images that meet the preset threshold are selected from all the captured images, and the camera parameters used for each target captured image are recorded, which are the target camera parameters.

[0134] Exemplarily, common features (such as specific focal length, aperture combination, etc.) can be extracted from the selected target camera parameters as the optimized basic settings. Additionally, additional tests can be performed on the selected combination of target camera parameters to ensure that good image quality can be maintained under different samples and environmental conditions. According to the test results, a set of optimal camera parameter combinations is selected, and combined with the camera recognition algorithm, the optimal camera recognition mode for the current sample is formed to ensure that various container models can be stably recognized in practical applications.

[0135] Optionally, the determination of the camera recognition mode corresponding to the container model based on the target camera parameters may specifically include: preprocessing the target captured image; performing edge recognition on the preprocessed target captured image through different vision algorithms; obtaining the target vision algorithm whose edge recognition result meets the requirements; and obtaining the camera recognition mode corresponding to the container model according to the target vision algorithm and the target camera parameters.

[0136] For example, the edges of the preprocessed target captured image can be recognized through methods such as Canny edge detection, Sobel operator, Laplacian operator, or deep learning methods. The key parameters (such as threshold, kernel size, etc.) of each algorithm are adjusted to obtain the best effect. Then, quantitative indicators such as intersection over union (IoU), F1 score, etc. are used to evaluate the edge detection results of each algorithm. Combining with the recognition accuracy requirements for the handling object, the algorithm that can stably detect all key features or has relatively high comprehensive computational efficiency is selected as the target vision algorithm. Furthermore, the camera recognition mode corresponding to the container model is determined by combining the target vision algorithm and the target camera parameters.

[0137] Optionally, the method of this embodiment may further include: sending the mapping relationship between the model identifiers of different container models and the camera recognition mode to the robot.

[0138] In some examples, once the camera recognition mode corresponding to a container model is confirmed to be valid, it can be integrated into the RMS system and become part of the optional task instructions. With the accumulation of more data and the development of technology, the recognition modes of each container type are continuously monitored and updated to maintain the advancement and adaptability of the system. The optimal camera recognition mode is selected for each container type, and a mapping relationship table is maintained in the database to record each container model identifier and its corresponding optimal camera recognition mode (including camera parameters and visual algorithm configuration). Special modes can also be saved specifically for different color container types (such as black, red, dark gray, navy blue, etc.). The mapping relationship between the model identifiers of different container models and the camera recognition mode is sent to the robot. The robot can directly determine the camera recognition mode based on the model identifier issued by RMS, greatly improving the recognition accuracy and work efficiency of the robot in actual operation.

[0139] Optionally, the method of this embodiment may also include: receiving a target camera recognition mode sent by the robot and used when the edge recognition of the transported object is successful; updating the mapping relationship between the model identifier of the container model and the camera recognition mode based on the target camera recognition mode; and sending the updated mapping relationship to the robot.

[0140] For example, if the robot fails to successfully identify the object to be transported according to the camera recognition mode indicated by the RMS, the recognition mode of the depth camera is adjusted again, such as gradually modifying the camera parameters based on the predefined adjustment strategy, or trying to use different edge detection algorithms or adjusting the parameters of the existing algorithm, and repeating the above adjustment and recognition process until the edge of the transported object is successfully identified. Once successfully identified, the currently used camera parameters and visual algorithm configuration are immediately saved as the target camera recognition mode, and the message is actively sent back to the RMS through the communication interface. The RMS can then associate the new target camera recognition mode with the corresponding container model identifier and update the mapping relationship between the camera recognition mode and the corresponding container model identifier in the database. Machine learning or reinforcement learning algorithms can also be introduced to enable the system to automatically optimize the recognition mode based on historical data and real-time feedback, further improve the recognition success rate and efficiency, and ensure that the container model recognition mode in the automated handling system is adjusted and optimized in a timely manner, thereby improving the accuracy and stability of the system.

[0141] Compared with the current related technologies, the embodiments of the present disclosure first obtain the appearance features of different handling objects; then configure container models corresponding to different handling objects according to the appearance features; and then determine the camera recognition modes corresponding to different container models, and determine the preset mapping rules between the model identifiers of different container models and different characters. By applying the method of this embodiment, according to the preset mapping rules between the model identifiers of the container models configured corresponding to different handling objects and different characters, the applicable camera recognition mode can be selected according to different container models, so that the robot can quickly and effectively identify the edge of the handling object, and make precise action responses accordingly, saving the time consumed for adjusting camera parameters and identifying the edge of the handling object, improving the efficiency of the robot in performing the handling task, and also enhancing the flexibility and robustness of the system.

[0142] Further, to illustrate the specific execution process on the robot side, Figure 5 The flowchart of an exception handling method according to an embodiment of the present disclosure is shown. Applied to the execution on the side of the warehousing robot, it may include the following steps.

[0143] Step 301, receive the handling instruction of the handling object.

[0144] Among them, the handling instruction carries the model identifier of the container model corresponding to the handling object, and the container model is determined according to the parsing result of the attribute information of the handling object.

[0145] Exemplarily, the handling instruction may include a box-taking instruction, a box-placing instruction, etc. After the robot receives the box-taking instruction, it executes the task of moving the handling object out of its current position; correspondingly, after the robot receives the box-placing instruction, it executes the task of moving the handling object to the specified target position and placing it safely and accurately.

[0146] In some examples, the robot receives the handling instruction sent by the RMS, which contains the container model identifier to be recognized. After the robot receives the instruction, it parses the container model identifier information therein.

[0147] Step 302, according to the model identifier, determine the camera recognition mode used by the robot to recognize the handling object.

[0148] In some examples, the model identification is used by the robot to determine the camera recognition mode used when identifying the transported object. Different camera recognition modes can be used to identify different container models. By selecting the camera recognition mode according to the decision logic corresponding to the model identification, the recognition of the transported object can be quickly performed. When the robot receives a task instruction, it loads the camera recognition mode for a specific container type from the RMS, which includes pre-adjusted camera hardware parameters (such as focal length, aperture, ISO, etc.), and applies these parameters to its built-in or external camera system to ensure that the best image quality can be obtained when photographing the container. For example, if the transported object is a non-black container, the general mode is used for identification, where the general mode is the default recognition mode applicable to most non-black container types; if the transported object is determined to be a black container according to the model identification, the recognition mode corresponding to the black container is selected, such as adjusting parameters and algorithms to relax motor adjustment, increase compensation values, etc.

[0149] In some examples, there are corresponding recognition modes for red cargo box type, dark gray cargo box type, navy blue cargo box type, etc.

[0150] Optionally, the method of this embodiment may further include: receiving a mapping relationship between model identifiers of different container models and camera recognition modes; accordingly, step 302 may specifically include: determining the camera recognition mode used by the robot to identify the transported object by querying the mapping relationship.

[0151] Exemplarily, RMS integrates the camera recognition mode corresponding to the container model and makes it part of the optional task instruction. With the accumulation of more data and the development of technology, the recognition mode of each container type is continuously monitored and updated to maintain the advancement and adaptability of the system. RMS creates the optimal camera recognition mode for each container type and maintains a mapping relationship table in the database to record each container model identifier and its corresponding optimal camera recognition mode (including camera parameters and visual algorithm configuration). It can also save special modes for different color container types (such as black, red, dark gray, navy blue, etc.). The robot receives the mapping relationship between the model identifier and the camera recognition mode of different container models. The robot can directly determine the camera recognition mode based on the model identifier issued by RMS, which greatly improves the recognition accuracy and work efficiency of the robot in actual operation.

[0152] Optionally, step 302 may specifically include: determining camera parameters and visual algorithms used by the robot to identify the transported object.

[0153] Exemplarily, the robot can directly determine the currently applicable camera recognition mode according to the mapping relationship between the model identifier of different container models and the camera recognition mode, or can also initiate a query request for the model identifier to the database through a preset communication interface, and the database returns the optimal camera recognition mode matching the model identifier.

[0154] Optionally, the method of this embodiment may further include: performing edge recognition on the handling object based on camera parameters and vision algorithms.

[0155] In some examples, the settings of the depth camera are adjusted according to the determined camera parameters, the specified vision algorithm and its parameter configuration are loaded, the handling object is photographed according to the configured parameters, and the robot performs recognition and edge detection on the handling object according to the camera parameters and vision algorithms configured in the camera recognition mode, and extracts key features.

[0156] Optionally, the above-mentioned edge recognition of the handling object based on camera parameters and vision algorithms may specifically include: photographing the handling object by using the robot camera device according to the camera parameters; performing edge recognition on the photographed image according to the vision algorithm to obtain an edge map marked with edge positions; matching the edge features in the edge map with the edge features of the container model corresponding to the model identifier; determining whether the edge recognition is successful according to the matching result, and obtaining the successfully recognized edge position in the case of successful edge recognition.

[0157] In some embodiments, the robot starts to perform edge detection based on image processing technology or a deep learning model according to the selected vision algorithm. After being processed by the algorithm, a new image marked with potential edge positions, that is, an edge map, is obtained. Meaningful information, such as straight line segments, curves, etc., is extracted from the edge map, and these information can help determine the specific contour of the cargo box. Then, using the pre-constructed container model, try to match the edge features in the image with the expected cargo box shape. If there are any uncertainties or the preliminary results are inaccurate, the container model can be rechecked and corrected according to the known cargo box size, relative position, etc.

[0158] Optionally, the method of this embodiment may further include: during the process of recognizing the handling object by using the camera recognition mode determined according to the model identifier, if the edge recognition fails after a preset number of attempts, the handling object is recognized by using the replaced camera recognition mode; sending the target camera recognition mode used when the edge recognition of the handling object is successful to update the mapping relationship between the model identifier of the container model and the camera recognition mode.

[0159] Exemplarily, after confirming the positions of the edges of the cargo box, these position information need to be converted into spatial coordinate points recognizable by the robot. Based on these coordinate points, a safe and efficient path is planned for the robot to ensure that the robotic arm or other grasping devices can accurately move to the predetermined positions and complete the operation of picking up the box. Subsequently, the robot will execute corresponding actions according to this predetermined path, thereby achieving precise box grasping. By applying the method of this embodiment, the robot can effectively identify the edges of the cargo box according to the applicable camera recognition mode and make precise action responses accordingly. This process not only improves work efficiency but also enhances the flexibility and robustness of the system.

[0160] In some examples, if the recognition of the handling object is successful according to the camera recognition mode determined based on the model identifier, the subsequent tasks are continued; if all attempts fail after the preset number of attempts, the recognition mode is readjusted (including adjusting camera parameters and vision algorithms) until success, and the finally used mode is fed back to the RMS. By recording each recognition operation, it can be used to evaluate the effect of the current camera recognition mode, and based on this accumulated data, continuously adjust and optimize the camera settings, vision algorithms and their parameters, and retrain the machine learning model if necessary to ensure that the system can adapt to different environmental conditions.

[0161] Compared with the current related technologies, the embodiments of the present disclosure first receive a handling instruction for a handling object, where the handling instruction carries a model identifier of a container model corresponding to the handling object, and the container model is determined according to the parsing result of the attribute information of the handling object; then, according to the model identifier, the camera recognition mode used by the robot to recognize the handling object is determined. By applying the method of this embodiment, for the handling instruction received by the robot, the container model corresponding to the handling object can be determined according to the model identifier in the handling instruction, and an applicable camera recognition mode can be selected to quickly and effectively identify the edges of the handling object and make precise action responses accordingly, saving the time consumed for adjusting camera parameters and recognizing the edges of the handling object, improving the efficiency of the robot in executing handling tasks, and enhancing the flexibility and robustness of the system.

[0162] Further, as Figures 1 to 4 a specific implementation of the example shown, the embodiments of the present disclosure provide a robot control device, which can be applied to the controller side, as Figure 6 shown, the device includes: an acquisition module 41 and a sending module 42.

[0163] The acquisition module 41 is configured to acquire the attribute information of the handling object; parse the attribute information and determine the container model corresponding to the handling object according to the parsing result; acquire the model identifier of the container model;

[0164] A sending module 42, configured to send a handling instruction of the handling object to the robot, where the handling instruction carries the model identifier, and the model identifier is used for the robot to determine the camera recognition mode used when recognizing the handling object.

[0165] In some embodiments, the obtaining module 41 is further configured to obtain the object identifier of the handling object from the parsing result; determine the model identifier according to the object identifier.

[0166] In some embodiments, the obtaining module 41 is further configured to extract a target character at a preset position of the object identifier;

[0167] Determine the model identifier corresponding to the target character according to the preset mapping rule between the character and the model identifier.

[0168] In some embodiments, the obtaining module 41 is further configured to obtain the appearance features of different handling objects; configure container models corresponding to different handling objects according to the appearance features; determine the camera recognition modes corresponding to different container models respectively, and determine the preset mapping rule between the model identifiers of different container models and different characters.

[0169] In some embodiments, the obtaining module 41 is further configured to obtain model parameters corresponding to the container model, where the model parameters include one or more of the size, color, material, and surface texture of the container model; determine camera parameters based on the model parameters and the preset working environment, where the camera parameters include one or more of the focal length, aperture size, sensitivity, exposure time, and white balance; perform test shooting on the model sample of the container model according to the camera parameters; determine the camera recognition mode corresponding to the container model according to the test shooting result.

[0170] In some embodiments, the obtaining module 41 is further configured to perform image quality analysis on the captured images of the model sample under different camera parameter conditions respectively; obtain the target camera parameters corresponding to the target captured images with image quality greater than the preset threshold; determine the camera recognition mode corresponding to the container model according to the target camera parameters.

[0171] In some embodiments, the obtaining module 41 is further configured to preprocess the target captured image; perform edge recognition on the preprocessed target captured image through different vision algorithms; obtain the target vision algorithm whose edge recognition result meets the requirements; obtain the camera recognition mode corresponding to the container model according to the target vision algorithm and the target camera parameters.

[0172] In some embodiments, the sending module 42 is further configured to send the mapping relationship between the model identifiers of different container models and the camera recognition modes to the robot.

[0173] In some embodiments, the sending module 42 is further configured to receive the target camera recognition mode used by the robot when the edge recognition of the handling object is successful; update the mapping relationship between the model identifier of the container model and the camera recognition mode according to the target camera recognition mode; and send the updated mapping relationship to the robot.

[0174] In some embodiments, the obtaining module 41 is further configured to perform interface verification according to the parsing result to determine whether there is a container model corresponding to the handling object; if there is no container model corresponding to the handling object, it is determined that adding the handling object to the database fails and an appropriate error message is triggered for output, where the objects successfully added to the database can be tracked, scheduled, and managed; if there is a container model corresponding to the handling object, the handling object is added to the database and the container model corresponding to the handling object is determined.

[0175] In some embodiments, the obtaining module 41 is further configured to obtain the object identifier of the handling object from the parsing result; determine whether the object identifier conforms to a preset identifier form, where the identifier of the preset identifier form has a corresponding container model; if the object identifier conforms to the preset identifier form, it is determined that there is a container model corresponding to the handling object; if the object identifier does not conform to the preset identifier form, it is determined that there is no container model corresponding to the handling object.

[0176] Further, as Figure 5 a specific implementation of the example shown, an embodiment of the present disclosure provides a robot control device, which can be applied to the side of a warehousing robot, such as Figure 7 shown, the device includes: a receiving module 51 and a determining module 52.

[0177] The receiving module 51 is configured to receive a handling instruction of a handling object, where the handling instruction carries a model identifier of a container model corresponding to the handling object, and the container model is determined according to a parsing result of attribute information of the handling object;

[0178] The determining module 52 is configured to determine a camera recognition mode used by the robot to recognize the handling object according to the model identifier.

[0179] In some embodiments, the determining module 52 is further configured to receive a mapping relationship between model identifiers of different container models and camera recognition modes; and determine a camera recognition mode used by the robot to recognize the handling object by querying the mapping relationship.

[0180] In some embodiments, the determination module 52 is further configured to determine the camera parameters and vision algorithms used by the robot when identifying the handling object.

[0181] In some embodiments, the determination module 52 is further configured to perform edge recognition on the handling object based on the camera parameters and vision algorithms.

[0182] In some embodiments, the determination module 52 is further configured to use the robot camera device to capture an image of the handling object according to the camera parameters; perform edge recognition on the captured image according to the vision algorithm to obtain an edge map marked with edge positions; match the edge features in the edge map with the edge features of the container model corresponding to the model identifier; determine whether the edge recognition is successful based on the matching result, and obtain the successfully recognized edge position in the case of successful edge recognition.

[0183] In some embodiments, the determination module 52 is further configured to, during the process of identifying the handling object using the camera recognition mode determined according to the model identifier, if the edge recognition fails after a preset number of attempts, use the replaced camera recognition mode to identify the handling object; send the target camera recognition mode used when the edge recognition of the handling object is successful to update the mapping relationship between the model identifier of the container model and the camera recognition mode.

[0184] It should be noted that for other corresponding descriptions of the various functional units involved in the robot control device provided in the embodiments of the present disclosure, reference can be made to Figures 1 to 5 the corresponding descriptions therein, which will not be elaborated here.

[0185] Based on the above example as Figures 1 to 5 shown, correspondingly, the embodiments of the present disclosure further provide a computer-readable storage medium, on which a computer program is stored, and when the computer program is executed by a processor, it implements the example method as Figures 1 to 5 shown above.

[0186] Based on the above example as Figures 1 to 5 shown, correspondingly, the embodiments of the present disclosure further provide a computer program product, including a computer program, and when the computer program is executed by a processor, it implements the example method as Figures 1 to 5 shown above.

[0187] Based on such an understanding, the technical solution of the present disclosure can be embodied in the form of a software product, and the software product can be stored in a non-volatile storage medium (which can be a CD-ROM, a USB flash drive, a mobile hard disk, etc.), including several instructions for causing a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods in various implementation scenarios of the present disclosure.

[0188] According to an embodiment of the present disclosure, the embodiments of the present disclosure further provide an electronic device, a readable storage medium, and a computer program product.

[0189] Figure 8 FIG. shows a schematic block diagram of an exemplary electronic device 1000 that may be used to implement embodiments of the present disclosure. The electronic device is intended to represent various forms of digital computers, such as, for example, laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device may also represent various forms of mobile devices, such as, for example, personal digital processors, cellular telephones, smart phones, wearable devices, and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely exemplary and are not intended to limit the implementation of the present disclosure described and / or claimed herein.

[0190] As Figure 8 shown, the device 1000 includes a computing unit 1001 that can perform various appropriate actions and processes according to a computer program stored in a ROM (Read-Only Memory) 1002 or a computer program loaded from a storage unit 1008 into a RAM (Random Access Memory) 1003. In the RAM 1003, various programs and data required for the operation of the device 1000 may also be stored. The computing unit 1001, the ROM 1002, and the RAM 1003 are connected to each other via a bus 1004. An I / O (Input / Output) interface 1005 is also connected to the bus 1004.

[0191] A plurality of components in the device 1000 are connected to the I / O interface 1005, including: an input unit 1006, such as a keyboard, a mouse, etc.; an output unit 1007, such as various types of displays, speakers, etc.; a storage unit 1008, such as a magnetic disk, an optical disk, etc.; and a communication unit 1009, such as a network card, a modem, a wireless communication transceiver, etc. The communication unit 1009 allows the device 1000 to exchange information / data with other devices via a computer network such as the Internet and / or various telecommunication networks.

[0192] The computing unit 1001 can be various general-purpose and / or special-purpose processing components with processing and computing capabilities. Some examples of the computing unit 1001 include, but are not limited to, CPU (Central Processing Unit), GPU (Graphic Processing Units), various dedicated AI (Artificial Intelligence) computing chips, various computing units running machine learning model algorithms, DSP (Digital Signal Processor), and any suitable processor, controller, microcontroller, etc. The computing unit 1001 executes the various methods and processes described above, such as the control method of the robot. For example, in some embodiments, the control method of the robot can be implemented as a computer software program, which is tangibly contained in a machine-readable medium, such as the storage unit 1008. In some embodiments, part or all of the computer program can be loaded and / or installed onto the device 1000 via the ROM 1002 and / or the communication unit 1009. When the computer program is loaded into the RAM 1003 and executed by the computing unit 1001, one or more steps of the methods described above can be executed. Alternatively, in other embodiments, the computing unit 1001 can be configured to execute the aforementioned control method of the robot in any other suitable manner (e.g., by means of firmware).

[0193] The various embodiments of the systems and techniques described above in this document can be implemented in digital electronic circuit systems, integrated circuit systems, FPGAs (Field Programmable Gate Arrays), ASICs (Application-Specific Integrated Circuits), ASSPs (Application Specific Standard Products), SoCs (System On Chip), CPLDs (Complex Programmable Logic Devices), computer hardware, firmware, software, and / or combinations thereof. These various embodiments can include: being implemented in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which can be a special or general programmable processor, receiving data and instructions from a storage system, at least one input device, and at least one output device, and transmitting the data and instructions to the storage system, the at least one input device, and the at least one output device.

[0194] The program code for implementing the methods of the present disclosure can be written in any combination of one or more programming languages. These program codes can be provided to a processor or controller of a general-purpose computer, a special-purpose computer, or other programmable data processing devices, such that when the program codes are executed by the processor or controller, the functions / operations specified in the flowcharts and / or block diagrams are implemented. The program codes can be executed entirely on the machine, partially on the machine, executed partially on the machine as an independent software package and partially on a remote machine, or executed entirely on a remote machine or server.

[0195] In the context of the present disclosure, a machine-readable medium can be a tangible medium that can contain or store a program for use by or in connection with an instruction execution system, apparatus, or device. A machine-readable medium can be a machine-readable signal medium or a machine-readable storage medium. A machine-readable medium can include, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, apparatus, or devices, or any suitable combination of the foregoing. More specific examples of machine-readable storage media would include electrical connections based on one or more wires, portable computer disks, hard disks, RAM, ROM, EPROM (Electrically Programmable Read-Only-Memory), or flash memory, optical fibers, CD-ROM (Compact Disc Read-Only Memory), optical storage devices, magnetic storage devices, or any suitable combination of the foregoing.

[0196] In order to provide interaction with a user, the systems and techniques described herein can be implemented on a computer having: a display device (e.g., a CRT (Cathode-Ray Tube) or an LCD (Liquid Crystal Display) monitor) for displaying information to the user; and a keyboard and a pointing device (e.g., a mouse or a trackball) through which the user can provide input to the computer. Other kinds of devices can also be used to provide interaction with the user; for example, the feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form (including acoustic input, voice input, or tactile input).

[0197] The systems and techniques described herein can be implemented in a computing system including backend components (e.g., as a data server), or a computing system including middleware components (e.g., an application server), or a computing system including frontend components (e.g., a user computer having a graphical user interface or a web browser through which a user can interact with an implementation of the systems and techniques described herein), or a computing system including any combination of such backend components, middleware components, or frontend components. The components of the system can be interconnected to each other by digital data communication in any form or medium (e.g., a communication network). Examples of communication networks include: LAN (Local Area Network), WAN (Wide Area Network), the Internet, and blockchain networks.

[0198] A computer system can include a client and a server. The client and the server are generally far from each other and usually interact through a communication network. The client-server relationship is created by computer programs running on respective computers and having a client-server relationship with each other. The server can be a cloud server, also known as a cloud computing server or a cloud host, which is a host product in the cloud computing service system, solving the defects of difficult management and weak business scalability existing in traditional physical hosts and VPS services ("Virtual Private Server", or simply "VPS"). The server can also be a server of a distributed system, or a server combined with blockchain.

[0199] Herein, it should be noted that artificial intelligence is a discipline that studies enabling a computer to simulate certain human thought processes and intelligent behaviors (such as learning, reasoning, thinking, planning, etc.), and it has both hardware-level technologies and software-level technologies. Artificial intelligence hardware technologies generally include technologies such as sensors, dedicated artificial intelligence chips, cloud computing, distributed storage, and big data processing; artificial intelligence software technologies mainly include several major directions such as computer vision technology, speech recognition technology, natural language processing technology, and machine learning / deep learning, big data processing technology, and knowledge graph technology.

[0200] It should be understood that various forms of processes shown above can be used, with steps reordered, added, or deleted. For example, the steps recited in this disclosure can be executed in parallel, sequentially, or in a different order, as long as the desired results of the technical solutions disclosed in this disclosure can be achieved, and no limitations are imposed herein.

[0201] The above specific embodiments do not constitute a limitation on the protection scope of the present disclosure. Those skilled in the art should understand that various modifications, combinations, sub - combinations and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions and improvements made within the spirit and principle of the present disclosure shall be included within the protection scope of the present disclosure.

Claims

1. A robot control method, characterized in that: include: Get the attribute information of the transported object; Parsing the attribute information, and determining a container model corresponding to the transport object according to the parsing result; Obtaining a model identifier of the container model; A transport instruction for the transport object is sent to the robot, wherein the transport instruction carries the model identifier, and the model identifier is used by the robot to determine a camera recognition mode used when identifying the transport object.

2. The method according to claim 1, characterized in that The step of determining the container model corresponding to the transport object according to the analysis result includes: Acquire the object identification of the transported object from the parsing result; The model identifier is determined according to the object identifier.

3. The method according to claim 2, characterized in that Determining the model identifier according to the object identifier includes: Extracting a target character at a preset position of the object identifier; According to a preset mapping rule between characters and model identifiers, a model identifier corresponding to the target character is determined.

4. The method according to claim 1, characterized in that Before obtaining the model identifier of the container model corresponding to the transport object, the method further includes: Obtain the appearance characteristics of different transport objects; Configuring container models corresponding to different transport objects according to the appearance features; Determine the camera recognition modes corresponding to different container models, and determine the preset mapping rules between the model identifiers of different container models and different characters.

5. The method according to claim 4, characterized in that Determine the camera recognition mode corresponding to the container model, including: Acquire model parameters corresponding to the container model, where the model parameters include one or more of size, color, material, and surface texture of the container model; Determine camera parameters based on the model parameters and a preset working environment, wherein the camera parameters include one or more of focal length, aperture size, sensitivity, exposure time, and white balance; Test shooting of a model sample of the container model according to the camera parameters; The camera recognition mode corresponding to the container model is determined based on the test shooting results.

6. The method according to claim 5, characterized in that The step of determining the camera recognition mode corresponding to the container model according to the test shooting result includes: Performing image quality analysis on images taken by the model samples under conditions of different camera parameters respectively; Obtain target camera parameters corresponding to target captured images whose image quality is greater than a preset threshold; A camera recognition mode corresponding to the container model is determined according to the target camera parameters.

7. The method according to claim 6, characterized in that Determining a camera recognition mode corresponding to the container model according to the target camera parameters includes: Preprocessing the target captured image; The pre-processed target images are subjected to edge recognition through different visual algorithms; Obtain edge recognition results that meet the requirements of the target visual algorithm; According to the target vision algorithm and the target camera parameters, a camera recognition mode corresponding to the container model is obtained.

8. The method according to claim 1, characterized in that The method further comprises: The mapping relationship between the model identifications of different container models and the camera recognition patterns is sent to the robot.

9. The method according to claim 8, characterized in that After sending the transport instruction of the transport object to the robot, the method further includes: receiving a target camera recognition mode sent by the robot used when the edge recognition of the transported object is successful; According to the target camera recognition mode, updating the mapping relationship between the model identifier of the container model and the camera recognition mode; Send the updated mapping relationship to the robot.

10. The method according to claim 1, characterized in that Before determining the container model corresponding to the transport object according to the analysis result, the method further includes: Performing interface verification according to the analysis result to determine whether there is a container model corresponding to the transport object; If there is no container model corresponding to the transport object, it is determined that the transport object fails to be added to the database and a corresponding error message is triggered to be output, wherein the object successfully added to the database can be tracked, scheduled and managed; The step of determining the container model corresponding to the transport object according to the analysis result includes: If there is a container model corresponding to the transport object, the transport object is added to the database and the container model corresponding to the transport object is determined.

11. The method according to claim 10, characterized in that Performing an interface check according to the analysis result to determine whether there is a container model corresponding to the transport object, including: Acquire the object identification of the transported object from the parsing result; Determining whether the object identifier conforms to a preset identifier format, wherein the identifier of the preset identifier format has a corresponding container model; If the object identification conforms to the preset identification format, it is determined that there is a container model corresponding to the transport object; If the object identification does not conform to the preset identification format, it is determined that there is no container model corresponding to the transport object.

12. A robot control method, characterized in that: include: receiving a transport instruction for a transport object, wherein the transport instruction carries a model identifier of a container model corresponding to the transport object, wherein the container model is determined based on a result of parsing attribute information of the transport object; A camera recognition mode used by the robot to identify the transport object is determined according to the model identifier.

13. The method according to claim 12, characterized in that Before determining, according to the model identifier, a camera recognition mode used by the robot when recognizing the transported object, the method further includes: Receive mapping relationships between model identifiers of different container models and camera recognition modes; Determining, according to the model identifier, a camera recognition mode used by the robot when identifying the transported object, including: The camera recognition mode used by the robot to identify the transported object is determined by querying the mapping relationship.

14. The method according to claim 12, characterized in that Determining, according to the model identifier, a camera recognition mode used by the robot when identifying the transported object, including: Determine the camera parameters and vision algorithm used by the robot to identify the transported object.

15. The method according to claim 14, characterized in that The method further comprises: The edge recognition of the transported object is performed based on the camera parameters and the visual algorithm.

16. The method according to claim 15, characterized in that Performing edge recognition on the transport object based on the camera parameters and the visual algorithm includes: Using a robot camera device to photograph the transported object according to the camera parameters; Perform edge recognition on the captured image according to the visual algorithm to obtain an edge map with edge positions marked; Matching edge features in the edge map with edge features of a container model corresponding to the model identifier; Determine whether edge recognition is successful based on the matching result, and obtain the position of the successfully recognized edge if edge recognition is successful.

17. The method according to claim 12, characterized in that The method further comprises: In the process of using the camera recognition mode determined according to the model identifier to identify the transported object, if edge recognition fails after a preset number of attempts, then using a replaced camera recognition mode to identify the transported object; The target camera recognition mode used when the edge recognition of the transported object is successful is sent to update the mapping relationship between the model identifier of the container model and the camera recognition mode.

18. A robot control device, characterized in that: include: An acquisition module, configured to acquire attribute information of a transported object; Parsing the attribute information, and determining the container model corresponding to the transport object according to the parsing result; acquiring the model identifier of the container model; The sending module is configured to send a transport instruction of the transport object to the robot, wherein the transport instruction carries the model identifier, and the model identifier is used by the robot to determine a camera recognition mode used when identifying the transport object.

19. A robot control device, characterized in that: include: A receiving module, configured to receive a transport instruction of a transport object, wherein the transport instruction carries a model identifier of a container model corresponding to the transport object, and the container model is determined based on a result of parsing attribute information of the transport object; The determination module is configured to determine, according to the model identifier, a camera recognition mode used by the robot when identifying the transport object.

20. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the method according to any one of claims 1 to 17 is implemented.

21. An electronic device comprising a storage medium, a processor, and a computer program stored in the storage medium and executable on the processor, characterized in that: When the processor executes the computer program, the method according to any one of claims 1 to 17 is implemented.

22. A computer program product comprising a computer program, characterized in that The computer program implements the method of any one of claims 1 to 17 when executed by a processor.

Images