A robot wireless communication system based on edge computing

Abstract

The present invention discloses a robot wireless communication system based on edge computing, which belongs to the field of production management technology, and specifically includes: storing multi-angle standard images of products; controlling a scanning robot to collect multi-angle images of finished products; identifying barcode information in the images, comparing standard images through edge computing, determining whether the finished products are abnormal or not, and determining the abnormal type and responsible process with a deep learning model; locating the actual production robot of the process where the abnormal product is located, and sending processing information; marking the processing information of the abnormal product as pending, querying unprocessed products with the same processing information, and producing robot detection prompts; if not found, repairing the actual production robot; the present invention realizes the intercommunication of production information and improves the collaborative efficiency between production robots.

Description

Technical Field

[0001] The present invention relates to the technical field of production management, and in particular to a robot wireless communication system based on edge computing. Background Art

[0002] In modern manufacturing, product quality control is crucial. Traditional quality inspection methods mainly rely on manual work, which is not only inefficient but also prone to errors. With the development of computer vision and deep learning technology, automated quality inspection has become possible. However, existing automated quality inspection systems usually need to transmit image data to a central server for processing, which causes data transmission delays and affects production efficiency.

[0003] In addition, the existing automated quality inspection system can usually only detect whether there are any abnormalities in the product, but cannot determine the responsible process for the abnormality. The various production robots cannot communicate with each other, and the collaborative efficiency is low. This is not enough to improve the production process and improve product quality.

[0004] Therefore, developing a robot wireless communication system based on edge computing that can detect product anomalies in real time and accurately and communicate information is of great significance for improving product quality and production efficiency. Summary of the invention

[0005] The purpose of the present invention is to provide a robot wireless communication system based on edge computing to solve the following technical problems:

[0006] Existing automated quality inspection systems can usually only detect whether there are any abnormalities in the product, but cannot determine the responsible process for the abnormality. There is no information exchange between the various production robots, and the collaborative efficiency is low.

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

[0008] A robot wireless communication system based on edge computing, comprising:

[0009] A database for storing standard images of processed products from several viewing angles;

[0010] A data acquisition module is used to control the finished product scanning robot to acquire the finished product image after the product is processed, and the finished product image also includes several viewing angles;

[0011] An image processing module is used to identify product processing information in the finished product image, where the product processing information is attached to the product surface via a barcode, compare the finished product image with the standard image via edge computing, determine whether the finished product has an abnormality, determine the abnormality type of the finished product via a deep learning model, and determine the responsible process for the abnormal product based on the abnormality type;

[0012] A data transmission module, used for locating the actual production robot corresponding to the responsible process of the abnormal product according to the product processing information, and sending the processing information of the abnormal product to the actual production robot;

[0013] The abnormality detection module is used to mark the processing information of abnormal products as pending information, query the unprocessed products with the same processing information as the pending information, and give a prompt when the production robot detects the unprocessed product; if the unprocessed product with the same processing information as the pending information is not found, the actual production robot is repaired.

[0014] As a further solution of the present invention: the barcode of the product processing information is set in a label, and the label is automatically affixed to the product by a label generating robot.

[0015] As a further solution of the present invention: the process of identifying abnormal finished products is as follows:

[0016] Perform grayscale conversion on the standard image and the finished image to generate a grayscale standard image and a grayscale finished image, obtain the grayscale value Gi(x,y) of any pixel in any grayscale standard image, where i represents the order of the grayscale standard image in all viewing angle grayscale standard images, (x,y) represents the coordinates of the pixel in the current grayscale image, and the grayscale value G0i(x,y) of the pixel with the same coordinates in the grayscale finished image of the same viewing angle of the grayscale standard image, calculate the difference between the grayscale values ​​of Gi(x,y) and G0i(x,y), if the difference is greater than a preset light change threshold, the pixel is an abnormal pixel, otherwise it is a normal pixel, traverse all standard images and finished images, if the proportion of abnormal pixels to all pixels is greater than a%, where a is a preset value, the finished product is abnormal.

[0017] As a further solution of the present invention: the process of determining the responsible process by the deep learning model is:

[0018] Collect abnormal images of abnormal products generated in each process of the production process, perform deep learning training on each process and the corresponding abnormal images, establish a deep learning model between the abnormal images and the processes, input the finished product images into the deep learning model, output the corresponding production process, and store the finished product images to update the model.

[0019] As a further solution of the present invention: the scanning robot collects finished product images from several viewing angles by changing the shooting angle of the camera to a preset angle.

[0020] As a further solution of the present invention: the data acquisition module sends the acquired finished product images to a database, and acquires standard images of products of corresponding models from the database for comparison.

[0021] As a further solution of the present invention: the data transmission module sends the processing information of the abnormal product to the actual production robot. After the actual production robot inspects all the products that have been produced, the finished products with the same pending information are sent to the scanning robot, and the scanning robot performs image capture on the finished products with the same pending information.

[0022] Beneficial effects of the present invention:

[0023] The present invention uses edge computing and deep learning models to analyze finished product images, thereby automatically identifying product processing information and any abnormalities. It can not only determine whether there are abnormalities in the finished product, but also determine the specific type of finished product abnormality and the responsible process that caused the abnormality through the deep learning model. It can mark the processing information of the abnormal product as pending, and query the database through wireless communication whether there are unprocessed products with the same processing information. When the production robot detects these unprocessed products, the system will issue a prompt to prevent potential quality problems from occurring, realize the intercommunication of production information, and improve the collaborative efficiency between production robots. BRIEF DESCRIPTION OF THE DRAWINGS

[0024] The present invention will be further described below in conjunction with the accompanying drawings.

[0025] Figure 1 It is a flow chart of a robot wireless communication system based on edge computing of the present invention. DETAILED DESCRIPTION

[0026] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0027] See also Figure 1 As shown, the present invention is a robot wireless communication system based on edge computing, comprising:

[0028] A database for storing standard images of processed products from several viewing angles;

[0029] A data acquisition module is used to control the finished product scanning robot to acquire the finished product image after the product is processed, and the finished product image also includes several viewing angles;

[0030] An image processing module is used to identify product processing information in the finished product image, where the product processing information is attached to the product surface via a barcode, compare the finished product image with the standard image via edge computing, determine whether the finished product has an abnormality, determine the abnormality type of the finished product via a deep learning model, and determine the responsible process for the abnormal product based on the abnormality type;

[0031] A data transmission module, used for locating the actual production robot corresponding to the responsible process of the abnormal product according to the product processing information, and sending the processing information of the abnormal product to the actual production robot;

[0032] The abnormality detection module is used to mark the processing information of abnormal products as pending information, query the unprocessed products with the same processing information as the pending information, and give a prompt when the production robot detects the unprocessed product; if the unprocessed product with the same processing information as the pending information is not found, the actual production robot is repaired.

[0033] The present invention achieves the following effects:

[0034] Real-time and high efficiency: Through the application of edge computing technology, the system can process and analyze data close to the production site. This not only reduces the time delay of data transmission to the central server, but also improves the efficiency of the entire detection process, ensuring that product quality control can be carried out in real time.

[0035] Accuracy and Automation: The system uses deep learning models to analyze finished product images to automatically identify product processing information and any anomalies. This approach significantly improves the accuracy and consistency of inspections compared to traditional manual inspections.

[0036] Abnormality type identification and responsibility tracing: The system can not only determine whether there are abnormalities in the finished product, but also determine the specific type of abnormality in the finished product and the responsible process that caused the abnormality through deep learning models. This helps manufacturers quickly locate the source of the problem and adjust the production process in a timely manner, thereby improving product quality.

[0037] Prevention and prompt function: The system can mark the processing information of abnormal products as pending and query the database to see if there are unprocessed products with the same processing information. When the production robot detects these unprocessed products, the system will issue a prompt to prevent potential quality problems.

[0038] Data-driven continuous improvement: By comparing standard images with finished product images and continuously updating the deep learning model, the system promotes a data-based continuous improvement process. Over time, the system will become more accurate and better able to adapt to changes on the production line.

[0039] Comprehensive cost savings: Automated anomaly detection and responsible process tracking reduce reliance on manual inspections and reduce labor costs. At the same time, by promptly discovering and correcting problems in the production process, material waste and rework costs are reduced, thereby reducing overall production costs.

[0040] Equipment maintenance and management: The system provides maintenance prompts for actual production robots. When it is found that there is no corresponding unprocessed product corresponding to abnormal information, it will recommend equipment maintenance, which helps to maintain the optimal operating state of production equipment and reduce production interruptions caused by equipment failure.

[0041] In another preferred embodiment of the present invention, the barcode of the product processing information is set in a label, and the label is automatically affixed to the product by a label generating robot.

[0042] In another preferred embodiment of the present invention, the process of identifying abnormal finished products is as follows:

[0043] Perform grayscale conversion on the standard image and the finished image to generate a grayscale standard image and a grayscale finished image, obtain the grayscale value Gi(x,y) of any pixel in any grayscale standard image, where i represents the order of the grayscale standard image in all viewing angle grayscale standard images, (x,y) represents the coordinates of the pixel in the current grayscale image, and the grayscale value G0i(x,y) of the pixel with the same coordinates in the grayscale finished image of the same viewing angle of the grayscale standard image, calculate the difference between the grayscale values ​​of Gi(x,y) and G0i(x,y), if the difference is greater than a preset light change threshold, the pixel is an abnormal pixel, otherwise it is a normal pixel, traverse all standard images and finished images, if the proportion of abnormal pixels to all pixels is greater than a%, where a is a preset value, the finished product is abnormal.

[0044] In another preferred embodiment of the present invention, the process of determining the responsible process by the deep learning model is as follows:

[0045] Collect abnormal images of abnormal products generated in each process of the production process, perform deep learning training on each process and the corresponding abnormal images, establish a deep learning model between the abnormal images and the processes, input the finished product images into the deep learning model, output the corresponding production process, and store the finished product images to update the model.

[0046] In another preferred embodiment of the present invention, the scanning robot collects finished product images from several viewing angles by changing the shooting angle of the camera to a preset angle.

[0047] In another preferred embodiment of the present invention, the data acquisition module sends the acquired finished product image to a database, and acquires a standard image of a corresponding model product from the database for comparison.

[0048] In another preferred embodiment of the present invention, the data transmission module sends the processing information of abnormal products to the actual production robot. After the actual production robot inspects all products that have been produced, the finished products with the same pending information are sent to the scanning robot, and the scanning robot performs image capture on the finished products with the same pending information.

[0049] The above is a detailed description of an embodiment of the present invention, but the content is only a preferred embodiment of the present invention and cannot be considered to limit the scope of implementation of the present invention. All equivalent changes and improvements made within the scope of the present invention should still fall within the scope of the patent coverage of the present invention.

Claims

1. A robot wireless communication system based on edge computing, characterized in that: include: A database for storing standard images of processed products from several viewing angles; A data acquisition module is used to control the finished product scanning robot to acquire the finished product image after the product is processed, and the finished product image also includes several viewing angles; An image processing module is used to identify product processing information in the finished product image, where the product processing information is attached to the product surface via a barcode, compare the finished product image with the standard image via edge computing, determine whether the finished product has an abnormality, determine the abnormality type of the finished product via a deep learning model, and determine the responsible process for the abnormal product based on the abnormality type; A data transmission module, used for locating the actual production robot corresponding to the responsible process of the abnormal product according to the product processing information, and sending the processing information of the abnormal product to the actual production robot; The abnormality detection module is used to mark the processing information of the abnormal product as pending information, query the unprocessed product with the same processing information as the pending information, and give a prompt when the production robot detects the unprocessed product; if the unprocessed product with the same processing information as the pending information is not found, the actual production robot is repaired; The process of identifying abnormal finished products is as follows: Perform grayscale conversion on the standard image and the finished image to generate a grayscale standard image and a grayscale finished image, obtain the grayscale value Gi(x,y) of any pixel in any grayscale standard image, where i represents the order of the grayscale standard image in all viewing angle grayscale standard images, (x,y) represents the coordinates of the pixel in the current grayscale image, and the grayscale value G0i(x,y) of the pixel with the same coordinates in the grayscale finished image of the same viewing angle of the grayscale standard image, calculate the difference between the grayscale values ​​of Gi(x,y) and G0i(x,y), if the difference is greater than a preset light change threshold, the pixel is an abnormal pixel, otherwise it is a normal pixel, traverse all standard images and finished images, if the proportion of abnormal pixels to all pixels is greater than a%, a is a preset value, the finished product is abnormal; The process of deep learning model judging responsible process is as follows: Collect abnormal images of abnormal products generated in each process of the production process, perform deep learning training on each process and the corresponding abnormal images, establish a deep learning model between the abnormal images and the processes, input the finished product images into the deep learning model, output the corresponding production process, and store the finished product images to update the model.

2. According to claim 1, a robot wireless communication system based on edge computing is characterized in that: The barcode of the product processing information is set in the label, and the label is automatically affixed to the product by a label generating robot.

3. The robot wireless communication system based on edge computing according to claim 1, characterized in that: The scanning robot collects finished product images from several viewing angles by changing the shooting angle of the camera to a preset angle.

4. The robot wireless communication system based on edge computing according to claim 1, characterized in that: The data acquisition module sends the acquired finished product images to a database, and acquires standard images of corresponding model products from the database for comparison.

5. The robot wireless communication system based on edge computing according to claim 1, characterized in that: The data transmission module sends the processing information of abnormal products to the actual production robot. After the actual production robot inspects all products that have been produced, the finished products with the same pending information are sent to the scanning robot, and the scanning robot collects images of the finished products with the same pending information.

Images