Action simulation precision evaluation system suitable for automatic clothing manufacturing robot

Abstract

The invention belongs to the field of automatic clothing manufacturing, relates to a precision evaluation analysis technology, is used for solving the problem that optimization decision analysis cannot be carried out according to precision abnormal data in the prior art, and particularly relates to an action simulation precision evaluation system suitable for an automatic clothing manufacturing robot, which comprises a pose monitoring module, a precision evaluation module and an optimization analysis module, the pose monitoring module, the precision evaluation module and the optimization analysis module are in communication connection in sequence; the pose monitoring module is used for monitoring and analyzing the action imitation precision of the automatic clothing making robot; video recording is performed on a monitored object when the monitored object runs, and the running process of the monitored object is decomposed into a plurality of unit processes according to clothing making action regulations; according to the method, the action simulation precision of the automatic clothing making robot can be monitored and analyzed, a monitored image and a standard image are compared by combining video decomposition and image processing technologies to obtain a coincidence coefficient, and the robot action precision degree of a key frame is fed back through the coincidence coefficient.

Description

Technical Field

[0001] The present invention belongs to the field of automatic garment making, and relates to precision evaluation and analysis technology. Specifically, it is an action imitation precision evaluation system applicable to automatic garment-making robots. Background Art

[0002] Automatic garment-making robots are intelligent devices integrating a variety of advanced technologies, mainly used for the automated production and personalized customization of clothing. Through technologies such as computer vision, machine learning, and natural language processing, they can autonomously complete processes such as clothing cutting, sewing, and fitting, greatly improving production efficiency and product quality.

[0003] The invention patent with the publication number CN117182975A discloses a robot motion precision stability evaluation system based on temperature compensation. This evaluation system is used to fit the robot's repeated motion precision curve under constant temperature conditions through temperature compensation when the test environment temperature is uncontrollable, and evaluate whether the robot's repeated motion precision stability meets the standard. However, this evaluation system cannot perform optimization decision analysis based on precision anomaly data, resulting in low efficiency in the robot's precision optimization process.

[0004] In view of the above technical problems, this application proposes a solution. Summary of the Invention

[0005] The purpose of the present invention is to provide an action imitation precision evaluation system applicable to automatic garment-making robots, which is used to solve the problem that the prior art cannot perform optimization decision analysis based on precision anomaly data.

[0006] The technical problem to be solved by the present invention is: how to provide an action imitation precision evaluation system applicable to automatic garment-making robots that can perform optimization decision analysis based on precision anomaly data.

[0007] The purpose of the present invention can be achieved through the following technical solutions:

[0008] An action imitation precision evaluation system applicable to automatic garment-making robots includes a pose monitoring module, a precision evaluation module, and an optimization analysis module. The pose monitoring module, the precision evaluation module, and the optimization analysis module are communicatively connected in sequence.

[0009] The pose monitoring module is used to monitor and analyze the action imitation precision of the automatic garment-making robot: mark the automatic garment-making robot as the monitoring object, record a video of the monitoring object when it is running, decompose the running process of the monitoring object into several unit processes according to the garment-making action regulations, each unit process corresponding to a complete garment-making process, obtain the coincidence coefficient of all key frames in the unit process and send it to the precision evaluation module.

[0010] The precision evaluation module is used to evaluate and analyze the action imitation precision of the automatic clothing-making robot: when the overall precision of the monitoring object does not meet the requirements, an optimization analysis signal is generated and sent to the optimization analysis module;

[0011] The optimization analysis module is used to optimize and analyze the action imitation accuracy of the automatic clothing-making robot: arrange the unit processes in the order of decomposition time from first to last to obtain a decomposition sequence, arrange the unit processes in the order of failure coefficient from small to large to obtain a failure sequence, mark the absolute value of the difference in the sequence numbers of the unit processes in the decomposition sequence and the failure sequence as the aging value of the unit process, sum and average the aging values of all unit processes to obtain the timing influence coefficient of the monitoring object, and analyze the optimization decision of the monitoring object through the timing influence coefficient.

[0012] Further, the process of obtaining the coincidence coefficient of the key frames in the unit process includes: decomposing the video recording according to the unit process to obtain corresponding unit sub-videos, decomposing the unit sub-videos into frames of images and extracting the key frames therein as monitoring images, retrieving the standard images corresponding to the key frames through the database, magnifying the monitoring images and the standard images into pixel grid images and performing gray-scale transformation, and then performing contour segmentation on the robots in the monitoring images and the standard images through the binary method to obtain the machine areas, and comparing the monitoring images and the standard images for coincidence to obtain the coincidence coefficient of the key frames.

[0013] Further, the specific process of comparing the monitoring image and the standard image for coincidence includes: marking the number of pixel grids occupied by the overlapping part of the machine areas in the monitoring image and the standard image as the coincidence value, marking the number of pixel grids occupied by the machine area in the standard image as the standard value, and marking the ratio of the coincidence value to the standard value as the coincidence coefficient of the monitoring image.

[0014] Further, the specific process of the precision evaluation module evaluating and analyzing the action imitation precision of the automatic clothing-making robot includes: comparing the coincidence coefficients of all key frames in the unit process with a preset coincidence threshold: if the coincidence coefficient is less than the coincidence threshold, mark the corresponding key frame as a failure frame; if the coincidence coefficient is greater than or equal to the coincidence threshold, mark the corresponding key frame as a valid frame; mark the ratio of the number of failure frames to the number of valid frames in the unit process as the failure coefficient of the unit process, and compare the failure coefficient with a preset failure threshold: if the failure coefficient is less than the failure threshold, determine that the action imitation precision of the monitoring object in the unit process meets the requirements, and mark the corresponding unit process as a precision qualified process; if the failure coefficient is greater than or equal to the failure threshold, determine that the action imitation precision of the monitoring object in the unit process does not meet the requirements, and mark the corresponding unit process as a precision unqualified process; evaluate the overall precision of the monitoring object during the operation process through the marking results of the precision unqualified processes.

[0015] Further, the specific process for evaluating the overall accuracy of the monitored object during operation includes: marking the ratio of the number of inaccurate process quantities to the number of unit processes during operation as the accuracy evaluation coefficient of the operation process, and comparing the accuracy evaluation coefficient with a preset accuracy evaluation threshold: if the accuracy evaluation coefficient is less than the accuracy evaluation threshold, it is determined that the overall accuracy of the monitored object during operation meets the requirements; if the accuracy evaluation coefficient is greater than or equal to the accuracy evaluation threshold, it is determined that the overall accuracy of the monitored object during operation does not meet the requirements.

[0016] Further, the specific process for performing optimization decision analysis on the monitored object includes: comparing the timing influence coefficient with a preset timing influence threshold: if the timing influence coefficient is less than the timing influence threshold, a fatigue optimization signal is generated and sent to the mobile terminal of the management personnel; if the timing influence coefficient is greater than or equal to the timing influence threshold, the execution sequence of the key frames in the unit process in the garment-making operation procedure is sorted to obtain an execution sequence, the sequence numbers of all failed frames in the execution sequence during operation form a failure set, the variance of the failure set is calculated to obtain a concentration coefficient, and the concentration coefficient is compared with a preset concentration threshold: if the concentration coefficient is less than the concentration threshold, the key frame corresponding to the element with the largest number of coincidences in the failure set is marked as the optimization link, a link optimization signal is generated and the link optimization signal and the optimization link are sent to the mobile terminal of the management personnel; if the concentration coefficient is greater than or equal to the concentration threshold, a device failure signal is generated and sent to the mobile terminal of the management personnel.

[0017] The present invention has the following beneficial effects:

[0018] 1. Through the pose monitoring module, the action imitation accuracy of the automatic garment-making robot can be monitored and analyzed. By combining video decomposition and image processing technologies, the coincidence coefficient is obtained by comparing the monitored image with the standard image, and the accuracy of the robot's action for the key frame is fed back through the coincidence coefficient;

[0019] 2. Through the accuracy evaluation module, the action imitation accuracy of the automatic garment-making robot can be evaluated and analyzed. The coincidence coefficients of the key frames of all unit processes during operation are processed to obtain an accuracy evaluation coefficient, and the overall accuracy is fed back through the accuracy evaluation coefficient. Moreover, when the overall accuracy is abnormal, an optimization analysis process is triggered to prevent the automatic garment-making robot from running for a long time in an irregular state;

[0020] 3. Through the optimization analysis module, the action imitation accuracy of the automatic garment-making robot can be optimized and analyzed. When the overall action accuracy of the monitored object does not meet the requirements, an optimization decision is automatically assigned to the monitored object to improve the optimization efficiency in abnormal states. Description of the Drawings

[0021] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings required for the description of the embodiments or the prior art. Obviously, the drawings in the following description are only some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can be obtained based on these drawings.

[0022] Figure 1 It is the system block diagram of the first embodiment of the present invention;

[0023] Figure 2 It is the method flowchart of the second embodiment of the present invention. Detailed implementation manners

[0024] The following will clearly and completely describe the technical solutions of the present invention in conjunction with the embodiments. Obviously, the described embodiments are only some embodiments of the present invention, rather than all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts belong to the scope of protection of the present invention.

[0025] Embodiment 1: As Figure 1 shown, a motion imitation accuracy evaluation system applicable to an automatic clothing-making robot includes a pose monitoring module, an accuracy evaluation module, and an optimization analysis module. The pose monitoring module, the accuracy evaluation module, and the optimization analysis module are communicatively connected in sequence.

[0026] The pose monitoring module is used to monitor and analyze the motion imitation accuracy of the automatic clothing-making robot: Mark the automatic clothing-making robot as the monitoring object, record a video of the monitoring object when it is running, decompose the running process of the monitoring object into several unit processes according to the clothing-making operation regulations, each unit process corresponding to a complete clothing-making process, decompose the video recording according to the unit process to obtain the corresponding unit sub-videos, decompose the unit sub-videos into frames of images and extract the key frames therein as monitoring images, retrieve the standard images corresponding to the key frames through the database, magnify the monitoring images and the standard images into pixel grid images and perform gray-scale transformation, and then perform contour segmentation on the robots in the monitoring images and the standard images through the binary method to obtain the machine areas. Overlap and compare the monitoring images and the standard images: Mark the number of pixel grids occupied by the overlapping part of the machine areas in the monitoring images and the standard images as the overlap value, mark the number of pixel grids occupied by the machine area in the standard image as the standard value, and mark the ratio of the overlap value to the standard value as the overlap coefficient of the monitoring image; Send the overlap coefficients of all the key frames of the unit process to the accuracy evaluation module.

[0027] The accuracy evaluation module is used to evaluate and analyze the action imitation accuracy of the automatic clothing-making robot: compare the coincidence coefficient of all key frames in the unit process with the preset coincidence threshold: if the coincidence coefficient is less than the coincidence threshold, mark the corresponding key frame as a failed frame; if the coincidence coefficient is greater than or equal to the coincidence threshold, mark the corresponding key frame as a valid frame; mark the ratio of the number of failed frames to the number of valid frames in the unit process as the failure coefficient of the unit process, and compare the failure coefficient with the preset failure threshold: if the failure coefficient is less than the failure threshold, it is determined that the action imitation accuracy of the monitored object in the unit process meets the requirements, and mark the corresponding unit process as a process with qualified accuracy; if the failure coefficient is greater than or equal to the failure threshold, it is determined that the action imitation accuracy of the monitored object in the unit process does not meet the requirements, and mark the corresponding unit process as a process with unqualified accuracy; mark the ratio of the number of processes with unqualified accuracy to the number of unit processes during operation as the accuracy evaluation coefficient during operation, and compare the accuracy evaluation coefficient with the preset accuracy evaluation threshold: if the accuracy evaluation coefficient is less than the accuracy evaluation threshold, it is determined that the overall accuracy of the monitored object during operation meets the requirements; if the accuracy evaluation coefficient is greater than or equal to the accuracy evaluation threshold, it is determined that the overall accuracy of the monitored object during operation does not meet the requirements, generate an optimization analysis signal and send the optimization analysis signal to the optimization analysis module.

[0028] The optimization analysis module is used to perform optimization analysis on the action imitation accuracy of the automatic clothing-making robot: arrange the unit processes in ascending order of decomposition time to obtain a decomposition sequence, arrange the unit processes in ascending order of failure coefficient to obtain a failure sequence, mark the absolute value of the difference in the sequence numbers of the unit processes in the decomposition sequence and the failure sequence as the timeliness value of the unit process, sum and average the timeliness values of all unit processes to obtain the timing influence coefficient of the monitored object, and compare the timing influence coefficient with the preset timing influence threshold: if the timing influence coefficient is less than the timing influence threshold, generate a fatigue optimization signal and send the fatigue optimization signal to the mobile terminal of the management personnel; if the timing influence coefficient is greater than or equal to the timing influence threshold, sort the execution order of the key frames in the unit process in the clothing-making action procedure to obtain an execution sequence, form a failure set from the sequence numbers of all failed frames in the operation process in the execution sequence, calculate the variance of the failure set to obtain a concentration coefficient, and compare the concentration coefficient with the preset concentration threshold: if the concentration coefficient is less than the concentration threshold, mark the key frame corresponding to the element with the largest number of coincidences in the failure set as the optimization link, generate a link optimization signal and send the link optimization signal and the optimization link to the mobile terminal of the management personnel; if the concentration coefficient is greater than or equal to the concentration threshold, generate a device failure signal and send the device failure signal to the mobile terminal of the management personnel.

[0029] Embodiment 2: As Figure 2As shown in the figure, a method for evaluating the action imitation accuracy applicable to an automatic clothing-making robot includes the following steps:

[0030] Step 1: Monitor and analyze the action imitation accuracy of the automatic clothing-making robot: Mark the automatic clothing-making robot as the monitoring object, decompose the running process when the monitoring object is running, and obtain the coincidence coefficient of each key frame in the unit process.

[0031] Step 2: Evaluate and analyze the action imitation accuracy of the automatic clothing-making robot: Mark the key frame as a failure frame or a valid frame according to the coincidence coefficient, mark the ratio of the number of failure frames to the number of valid frames in the unit process as the failure coefficient of the unit process, mark the unit process as a process with qualified accuracy or a process with unqualified accuracy through the failure coefficient, mark the ratio of the number of processes with unqualified accuracy in the running process to the number of unit processes as the accuracy evaluation coefficient of the running process, and determine whether the overall accuracy meets the requirements through the accuracy evaluation coefficient.

[0032] Step 3: Conduct an optimization analysis on the action imitation accuracy of the automatic clothing-making robot and generate a fatigue optimization signal, a link optimization signal, or a device failure signal.

[0033] An action imitation accuracy evaluation system applicable to an automatic clothing-making robot, when working, marks the automatic clothing-making robot as the monitoring object, decomposes the running process when the monitoring object is running, and obtains the coincidence coefficient of each key frame in the unit process; marks the key frame as a failure frame or a valid frame according to the coincidence coefficient, marks the ratio of the number of failure frames to the number of valid frames in the unit process as the failure coefficient of the unit process, marks the unit process as a process with qualified accuracy or a process with unqualified accuracy through the failure coefficient, marks the ratio of the number of processes with unqualified accuracy in the running process to the number of unit processes as the accuracy evaluation coefficient of the running process, and determines whether the overall accuracy meets the requirements through the accuracy evaluation coefficient; conducts an optimization analysis on the action imitation accuracy of the automatic clothing-making robot and generates a fatigue optimization signal, a link optimization signal, or a device failure signal.

[0034] The above content is only an example and explanation of the structure of the present invention. Those skilled in the art of this technology can make various modifications or supplements to the described specific embodiments or use similar methods for substitution. As long as they do not deviate from the structure of the invention or exceed the scope defined by this claim book, they should fall within the protection scope of the present invention.

[0035] In the description of this specification, the descriptions referring to the terms "one embodiment", "example", "specific example", etc. mean that the specific features, structures, materials or characteristics described in connection with the embodiment or example are included in at least one embodiment or example of the present invention. In this specification, the schematic representations of the above terms do not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described may be combined in any one or more embodiments or examples in a suitable manner.

[0036] The preferred embodiments of the present invention disclosed above are only used to help explain the present invention. The preferred embodiments do not describe all the details in detail, nor do they limit the present invention to only the specific embodiments. Obviously, many modifications and variations can be made according to the content of this specification. These embodiments are selected and specifically described in this specification in order to better explain the principles and practical applications of the present invention, so that those skilled in the art can well understand and utilize the present invention. The present invention is only limited by the claims and their full scope and equivalents.

Claims

1. An action imitation accuracy evaluation system applicable to an automatic clothing-making robot, characterized in that, It includes a pose monitoring module, an accuracy evaluation module, and an optimization analysis module. The pose monitoring module, the accuracy evaluation module, and the optimization analysis module are communicatively connected in sequence; The pose monitoring module is used to monitor and analyze the action imitation accuracy of the automatic clothing-making robot: Mark the automatic clothing-making robot as the monitoring object. When the monitoring object is running, video record the monitoring object. Decompose the running process of the monitoring object into several unit processes according to the clothing-making action regulations. Each unit process corresponds to a complete clothing-making process. Obtain the coincidence coefficient of all key frames in the unit process and send it to the accuracy evaluation module; The accuracy evaluation module is used to evaluate and analyze the action imitation accuracy of the automatic clothing-making robot: When the overall accuracy of the monitoring object does not meet the requirements, generate an optimization analysis signal and send the optimization analysis signal to the optimization analysis module; The optimization analysis module is used to optimize and analyze the action imitation accuracy of the automatic clothing-making robot: Arrange the unit processes in the order of decomposition time from first to last to obtain a decomposition sequence. Arrange the unit processes in the order of failure coefficient from small to large to obtain a failure sequence. Mark the absolute value of the difference in the sequence numbers of the unit processes in the decomposition sequence and the failure sequence as the aging value of the unit process. Sum and average the aging values of all unit processes to obtain the timing influence coefficient of the monitoring object. Analyze the optimization decision of the monitoring object through the timing influence coefficient.

2. The action imitation accuracy evaluation system for an automatic clothing-making robot according to claim 1, characterized in that, The process of obtaining the coincidence coefficient of the key frames in the unit process includes: Decompose the video recording according to the unit process to obtain the corresponding unit sub-video. Decompose the unit sub-video into frames of images and extract the key frames therein as monitoring images. Retrieve the standard images corresponding to the key frames through the database. Enlarge the monitoring images and the standard images into pixel grid images and perform gray-scale transformation. Then, use the binary method to segment the contours of the robots in the monitoring images and the standard images to obtain the machine areas. Compare the monitoring images and the standard images for coincidence to obtain the coincidence coefficient of the key frames.

3. The action imitation accuracy evaluation system for an automatic clothing-making robot according to claim 2, characterized in that, The specific process of comparing the monitoring images and the standard images for coincidence includes: Mark the number of pixel grids occupied by the overlapping part of the machine areas in the monitoring images and the standard images as the coincidence value. Mark the number of pixel grids occupied by the machine area in the standard image as the standard value. Mark the ratio of the coincidence value to the standard value as the coincidence coefficient of the monitoring image.

4. The action imitation accuracy evaluation system for an automatic clothing-making robot according to claim 3, characterized in that, The specific process of the accuracy evaluation module for evaluating and analyzing the action imitation accuracy of the automatic clothing robot includes: comparing the coincidence coefficient of all key frames in the unit process with the preset coincidence threshold: if the coincidence coefficient is less than the coincidence threshold, the corresponding key frame is marked as a failed frame; if the coincidence coefficient is greater than or equal to the coincidence threshold, the corresponding key frame is marked as a valid frame; marking the ratio of the number of failed frames to the number of valid frames in the unit process as the failure coefficient of the unit process, and comparing the failure coefficient with the preset failure threshold: if the failure coefficient is less than the failure threshold, it is determined that the action imitation accuracy of the monitored object in the unit process meets the requirements, and the corresponding unit process is marked as a process with qualified accuracy; if the failure coefficient is greater than or equal to the failure threshold, it is determined that the action imitation accuracy of the monitored object in the unit process does not meet the requirements, and the corresponding unit process is marked as a process with unqualified accuracy; evaluating the overall accuracy of the monitored object during the operation process through the marking result of the process with unqualified accuracy.

5. The action imitation accuracy evaluation system for an automatic clothing-making robot according to claim 4, characterized in that, The specific process of evaluating the overall accuracy of the monitored object during the operation process includes: marking the ratio of the number of processes with unqualified accuracy in the operation process to the number of unit processes as the accuracy evaluation coefficient of the operation process, and comparing the accuracy evaluation coefficient with the preset accuracy evaluation threshold: if the accuracy evaluation coefficient is less than the accuracy evaluation threshold, it is determined that the overall accuracy of the monitored object during the operation process meets the requirements; if the accuracy evaluation coefficient is greater than or equal to the accuracy evaluation threshold, it is determined that the overall accuracy of the monitored object during the operation process does not meet the requirements.

6. The action imitation accuracy evaluation system for an automatic clothing-making robot according to claim 5, characterized in that The specific process of optimizing decision analysis for the monitored object includes: comparing the timing influence coefficient with the preset timing influence threshold: if the timing influence coefficient is less than the timing influence threshold, a fatigue optimization signal is generated and sent to the mobile terminal of the management personnel; if the timing influence coefficient is greater than or equal to the timing influence threshold, the execution sequence of the key frames in the clothing making action regulations in the unit process is sorted to obtain the execution sequence, the set of serial numbers of all failed frames in the execution sequence during the operation process constitutes the failure set, the variance of the failure set is calculated to obtain the concentration coefficient, and the concentration coefficient is compared with the preset concentration threshold: if the concentration coefficient is less than the concentration threshold, the key frame corresponding to the element with the largest number of coincidences in the failure set is marked as the optimization link, a link optimization signal is generated and the link optimization signal and the optimization link are sent to the mobile terminal of the management personnel; if the concentration coefficient is greater than or equal to the concentration threshold, a device failure signal is generated and sent to the mobile terminal of the management personnel.

Images