A machine vision-based industrial part detection method and detection system

Abstract

This invention relates to the field of machine vision technology and provides a machine vision-based method and system for inspecting industrial parts. The invention involves initial image capture to divide the installation structure into multiple surfaces; acquiring disassembly monitoring data and collecting screw parts during the disassembly process; performing surface positioning detection on the disassembly monitoring data; generating and displaying multiple corresponding surface structure contours; and positioning and placing multiple screw parts according to their surface location. When appearance abnormalities are observed, replacements are made. This invention enables the application of machine vision to the inspection of industrial parts undergoing rework. By detecting and acquiring surface positioning data for multiple screw parts, generating and displaying multiple surface structure contours, and positioning and placing multiple screw parts according to their surface location, it provides accurate reference for manual assembly operations, effectively avoiding confusion regarding the position and specifications of screw parts during manual assembly, improving disassembly and assembly efficiency, and reducing damage to industrial products caused by disassembly and assembly.

Description

Technical Field

[0001] This invention belongs to the field of machine vision technology, and in particular relates to a machine vision-based method and system for inspecting industrial parts. Background Technology

[0002] Machine vision is a comprehensive technology that utilizes computer vision, image processing, and artificial intelligence to enable machines to recognize, understand, and judge visually. Its core is to acquire image data of target objects through optical imaging, input the data into a computer or embedded processing unit via an image acquisition card, and then use algorithms to analyze, recognize, and measure the images, thereby realizing functions such as object detection, positioning, recognition, sorting, measurement, and control.

[0003] Compared to human vision, machine vision has the advantages of high speed, high precision, strong stability and adaptability to harsh environments, and is widely used in industrial automation, intelligent manufacturing, robot navigation, quality inspection, security monitoring, medical imaging and other fields.

[0004] In the current technology, machine vision is not used in the rework process of industrial products. The disassembly and assembly operations are usually carried out manually. For some complex industrial products, there are often a large number and types of screw parts. When disassembling and assembling manually, the position and specifications of the screw parts may be confused, and errors may occur during the assembly process, requiring repeated attempts. The disassembly and assembly efficiency is low and it is easy to damage the industrial products. Summary of the Invention

[0005] The purpose of this invention is to provide a machine vision-based method and system for inspecting industrial parts, aiming to solve the technical problems existing in the prior art mentioned in the background.

[0006] The embodiments of the present invention are implemented as follows:

[0007] A machine vision-based method for inspecting industrial parts, the method specifically includes the following steps:

[0008] Initialize the shooting, identify the target industrial product, match the three-dimensional structural model of the target industrial product, identify the structural installation, and divide multiple installation structural surfaces;

[0009] The manual disassembly process of the target industrial product is monitored in real time to obtain disassembly monitoring data, and screws and parts collected during the disassembly process are collected.

[0010] Based on multiple installation structure surfaces, the disassembly monitoring data is subjected to surface positioning detection to obtain surface positioning data of multiple screw parts;

[0011] Based on multiple installation structural surfaces, multiple corresponding surface structure contours are generated and displayed. Based on multiple facet positioning data, multiple screw parts are facet positioning and placement on the multiple surface structure contours.

[0012] Perform visual inspection on multiple screw parts that are positioned in a faceted manner to determine if there are any appearance abnormalities, and replace them accordingly if any are found.

[0013] As a further limitation of the technical solution of this embodiment of the invention, the steps of initializing the shooting, identifying the target industrial product, matching the three-dimensional structural model of the target industrial product, and identifying the structural installation and dividing multiple installation structural surfaces specifically include the following steps:

[0014] Perform initial shooting and acquire initial shooting data;

[0015] Target identification is performed on the initial captured data to determine the target industrial product;

[0016] In a pre-set model database, match the three-dimensional structural model of the target industrial product.

[0017] Perform plane identification on the three-dimensional structural model to determine multiple model structural planes;

[0018] Structural installation identification is performed on multiple model structural planes, and multiple installation structural planes with screw assembly are selected from the multiple model structural planes.

[0019] As a further limitation of the technical solution of this invention embodiment, the real-time monitoring of the manual disassembly process of the target industrial product, the acquisition of disassembly monitoring data, and the collection of screw parts during the disassembly process specifically include the following steps:

[0020] Real-time monitoring of the manual disassembly process of target industrial products to obtain disassembly monitoring data;

[0021] Generate acceptance set instructions;

[0022] According to the receiving and collection instructions, open the receiving plate and use the receiving plate to receive and collect the screws and parts during the disassembly process.

[0023] As a further limitation of the technical solution of this embodiment of the invention, the step of performing facet positioning detection on the disassembly monitoring data based on multiple installation structure surfaces to obtain facet positioning data of multiple screw parts specifically includes the following steps:

[0024] The disassembly monitoring data is subjected to facet detection to determine the facet relationship between multiple screw parts and the mounting structure surface;

[0025] The disassembly monitoring data is used to perform positioning detection to determine the planar positions of multiple screw parts on the corresponding mounting structure surface;

[0026] The disassembly monitoring data is used to identify the order of installation and determine the installation order of the multiple installation surfaces.

[0027] By combining multiple facet relationships, multiple planar positions, and multiple structural installation sequences, facet positioning data for multiple screw parts is generated.

[0028] As a further limitation of the technical solution of this embodiment of the invention, the step of generating and displaying multiple corresponding surface structure contours according to multiple installation structure surfaces, and positioning and placing multiple screw parts on the multiple surface structure contours according to multiple facet positioning data, specifically includes the following steps:

[0029] Project the three-dimensional structural model according to the multiple installation structural surfaces to generate surface structural contours corresponding to the multiple installation structural surfaces;

[0030] Based on the multiple facet positioning data, plan the display positions of the multiple facet structure contours;

[0031] According to the multiple display positions, the multiple surface structure outlines are distinguished and displayed on the receiving plate;

[0032] Based on the multiple facet positioning data, plan the corresponding facet placement positions on the multiple facet structure contours;

[0033] According to the multiple faceted placement positions, multiple screw parts are positioned and placed in faceted positions.

[0034] As a further limitation of the technical solution of this invention embodiment, the step of performing visual inspection on multiple screw parts placed in faceted positioning to determine whether there are any visual abnormalities, and replacing them accordingly when there are visual abnormalities, specifically includes the following steps:

[0035] Multiple screw parts with faceted positioning are inspected and photographed to obtain inspection and photographic data;

[0036] Obtain standard images of multiple screw parts;

[0037] Based on multiple standard images, the detection and capture data are compared and detected to determine whether there are any appearance abnormalities.

[0038] When there is an appearance abnormality, select the abnormal screw from the plurality of screw parts;

[0039] Match the replacement screw to the abnormal screw;

[0040] Perform replacement placement control, remove the abnormal screw, and replace it with the replacement screw.

[0041] A machine vision-based industrial parts inspection system includes a structure installation recognition module, a real-time monitoring and processing module, a faceted positioning detection module, a faceted positioning placement module, and an abnormal placement replacement module, wherein:

[0042] The structural installation recognition module is used to perform initial shooting, identify the target industrial product, match the three-dimensional structural model of the target industrial product, and perform structural installation recognition, dividing multiple installation structural surfaces.

[0043] The real-time monitoring and processing module is used to monitor the manual disassembly process of the target industrial product in real time, acquire disassembly monitoring data, and collect screws and parts during the disassembly process.

[0044] The facet positioning detection module is used to perform facet positioning detection on the disassembly monitoring data based on multiple installation structure faces, and to obtain facet positioning data of multiple screw parts.

[0045] The faceted positioning and placement module is used to generate and display multiple corresponding faceted structure contours according to multiple installation structure surfaces, and to perform faceted positioning and placement of multiple screw parts on the multiple faceted structure contours according to multiple faceted positioning data.

[0046] The abnormal replacement placement module is used to perform visual inspection on multiple screw parts that are positioned faceted, determine whether there are any visual abnormalities, and replace them accordingly when there are visual abnormalities.

[0047] As a further limitation of the technical solution of this embodiment of the invention, the structure installation identification module specifically includes:

[0048] Initialize the shooting unit to perform initial shooting and acquire initial shooting data;

[0049] A target recognition unit is used to perform target recognition on the initial captured data to determine the target industrial product;

[0050] The model matching unit is used to match the three-dimensional structural model of the target industrial product in a preset model database.

[0051] A plane recognition unit is used to perform plane recognition on the three-dimensional structural model and determine multiple model structural planes;

[0052] The structural surface selection unit is used to identify the structural installation of multiple model structural planes and select multiple installation structural planes with screw assembly from the multiple model structural planes.

[0053] As a further limitation of the technical solution of this embodiment of the invention, the facet positioning and detection module specifically includes:

[0054] The facet detection unit is used to perform facet detection on the disassembly monitoring data to determine the facet relationship between multiple screw parts and the mounting structure surface.

[0055] The positioning detection unit is used to perform positioning detection on the disassembly monitoring data to determine the planar position of multiple screw parts on the corresponding mounting structure surface;

[0056] The order identification unit is used to identify the order of the disassembly monitoring data and determine the structural installation order corresponding to the multiple installation structural surfaces.

[0057] The facet positioning data generation unit is used to integrate multiple facet relationships, multiple planar positions, and multiple structural installation sequences to generate facet positioning data for multiple screw parts.

[0058] As a further limitation of the technical solution of this embodiment of the invention, the abnormal replacement placement module specifically includes:

[0059] The detection and imaging unit is used to detect and photograph multiple screw parts that are positioned in a faceted manner, and to acquire detection and imaging data.

[0060] A standard image acquisition unit is used to acquire standard images of multiple screw parts;

[0061] The comparison detection unit is used to compare and detect the detection data based on multiple standard images to determine whether there are any appearance abnormalities.

[0062] An abnormality selection unit is used to select an abnormal screw from a plurality of screw parts when there is an appearance abnormality;

[0063] A replacement matching unit is used to match the replacement screw corresponding to the abnormal screw;

[0064] The replacement placement control unit is used to perform replacement placement control, remove the abnormal screw, and replace the replacement screw.

[0065] Compared with the prior art, the beneficial effects of the present invention are:

[0066] (1) The present invention can apply machine vision to the inspection of industrial parts for the repair of industrial products, detect and obtain the facet positioning data of multiple screw parts, generate and display the facet structure contours, and place multiple screw parts in facet positioning, thereby providing accurate reference for manual assembly operations, effectively avoiding confusion of the position and specifications of screw parts during manual assembly, improving disassembly and assembly efficiency, and reducing damage to industrial products caused by disassembly and assembly.

[0067] (2) The present invention can detect and photograph multiple screw parts that are positioned in a faceted manner, and compare them with the standard images corresponding to the multiple screw parts to determine whether there is an appearance abnormality. Then, when there is an appearance abnormality, the corresponding replacement can be carried out to realize the automatic replacement of abnormal screws and further improve the disassembly and assembly efficiency of industrial product repair. Attached Figure Description

[0068] Figure 1 A flowchart of the machine vision-based industrial parts inspection method provided in an embodiment of the present invention is shown;

[0069] Figure 2 The diagram illustrates the application architecture of the machine vision-based industrial parts inspection system provided in this embodiment of the invention. Detailed Implementation

[0070] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention.

[0071] Understandably, in the current technology, machine vision is not used in the rework process of industrial products. The disassembly and assembly operations are usually carried out manually. For some complex industrial products, there are often a large number and types of screw parts. When disassembling and assembling manually, the position and specifications of the screw parts may be confused, and errors may occur during the assembly process, requiring repeated attempts. This results in low disassembly and assembly efficiency and is likely to cause damage to industrial products.

[0072] To address the aforementioned problems, this invention discloses a machine vision-based industrial parts inspection method and system. The method involves initial image capture to identify the target industrial product, matching its 3D structural model, and identifying the structural installation, dividing the product into multiple installation surfaces. Real-time monitoring of the manual disassembly process is conducted to acquire disassembly monitoring data, and the screws and other parts collected during disassembly are collected. Based on the multiple installation surfaces, the disassembly monitoring data is used for surface positioning detection to acquire surface positioning data for multiple screw parts. Multiple corresponding surface structure contours are generated and displayed according to the multiple installation surfaces, and multiple screw parts are positioned and placed on these contours according to the surface positioning data. The multiple screw parts are then subjected to appearance inspection to determine if any appearance abnormalities exist, and if any abnormalities are found, they are replaced accordingly. Machine vision can be applied to the inspection of industrial parts in the rework of industrial products. It can detect and acquire the facet positioning data of multiple screw parts, generate and display the contours of multiple facet structures, and position and place multiple screw parts facet by facet. This provides an accurate reference for manual assembly operations, effectively avoids confusion about the position and specifications of screw parts during manual assembly, improves disassembly and assembly efficiency, and reduces damage to industrial products caused by disassembly and assembly.

[0073] Specifically, Figure 1 A flowchart of a machine vision-based industrial parts inspection method provided by an embodiment of the present invention is shown.

[0074] In a preferred embodiment of the present invention, an industrial parts inspection method based on machine vision is provided, the method specifically including the following steps:

[0075] Step S101: Perform initial shooting, identify the target industrial product, match the three-dimensional structural model of the target industrial product, identify the structural installation, and divide multiple installation structural surfaces.

[0076] In this embodiment of the invention, before performing manual disassembly and assembly operations for product repair, the product to be disassembled and assembled needs to be placed on the disassembly and assembly workbench. Initial shooting is performed on the disassembly and assembly workbench to obtain initial shooting data. Based on preset product feature data, the initial shooting data is used for target feature recognition and comparative analysis to determine the target industrial product. Then, in the preset model database, the three-dimensional structural model of the target industrial product is matched. By automatically rotating the three-dimensional structural model at multiple angles, plane recognition is performed on the three-dimensional structural model at different rotation angles to determine multiple model structural planes on the surface of the three-dimensional structural model. Then, structural installation recognition is performed on the multiple model structural planes, and multiple installation structural surfaces with screw assembly are selected from the multiple model structural planes.

[0077] Specifically, in another preferred embodiment provided by the present invention, the steps of initializing the image, identifying the target industrial product, matching the three-dimensional structural model of the target industrial product, and identifying the structural installation and dividing multiple installation structural surfaces specifically include the following steps:

[0078] Perform initial shooting and acquire initial shooting data;

[0079] Target identification is performed on the initial captured data to determine the target industrial product;

[0080] In a pre-set model database, match the three-dimensional structural model of the target industrial product.

[0081] Perform plane identification on the three-dimensional structural model to determine multiple model structural planes;

[0082] Structural installation identification is performed on multiple model structural planes, and multiple installation structural planes with screw assembly are selected from the multiple model structural planes.

[0083] Furthermore, the machine vision-based industrial parts inspection method also includes the following steps:

[0084] Step S102: Monitor the manual disassembly process of the target industrial product in real time, obtain disassembly monitoring data, and collect screws and parts during the disassembly process.

[0085] In this embodiment of the invention, when workers are performing manual disassembly and assembly operations on target industrial products for product repair, the manual disassembly process of the target industrial products is monitored in real time to obtain disassembly monitoring data. At the same time, a receiving and collection instruction is generated. According to the receiving and collection instruction, the receiving plate is opened, and the screws and parts in the disassembly process of the target industrial products are received and collected through the receiving plate.

[0086] Specifically, in another preferred embodiment provided by the present invention, the real-time monitoring of the manual disassembly process of the target industrial product, the acquisition of disassembly monitoring data, and the collection of screw parts during the disassembly process specifically include the following steps:

[0087] Real-time monitoring of the manual disassembly process of target industrial products to obtain disassembly monitoring data;

[0088] Generate acceptance set instructions;

[0089] According to the receiving and collection instructions, open the receiving plate and use the receiving plate to receive and collect the screws and parts during the disassembly process.

[0090] Furthermore, the machine vision-based industrial parts inspection method also includes the following steps:

[0091] Step S103: Based on the multiple installation structure surfaces, perform surface positioning detection on the disassembly monitoring data to obtain surface positioning data of the multiple screw parts.

[0092] In this embodiment of the invention, by performing facet detection on the disassembly monitoring data, the facet relationships between multiple screw parts and multiple mounting structure surfaces are determined. Furthermore, by performing positioning detection on the disassembly monitoring data, the planar positions of multiple screw parts on the corresponding mounting structure surfaces are determined. Simultaneously, by performing sequence identification on the disassembly monitoring data, the structural disassembly sequence corresponding to multiple mounting structure surfaces during the disassembly of the target industrial product is determined. The structural disassembly sequences corresponding to multiple mounting structure surfaces are then reversed to obtain the structural installation sequence corresponding to multiple mounting structure surfaces. Finally, the multiple facet relationships, multiple planar positions, and multiple structural installation sequences are comprehensively organized to generate facet positioning data for multiple screw parts.

[0093] Understandably, the facet positioning data of multiple screw parts records the facet relationship, planar position, and structural installation order of each screw part. The facet relationship is the correspondence between the screw part and the mounting structure surface; the planar position is the specific position of the screw part on the corresponding mounting structure surface; and the structural installation order is the order of the mounting structure surface corresponding to the screw part among multiple mounting structure surfaces.

[0094] Specifically, in another preferred embodiment provided by the present invention, the step of performing facet positioning detection on the disassembly monitoring data based on multiple mounting structure surfaces to obtain facet positioning data of multiple screw parts specifically includes the following steps:

[0095] The disassembly monitoring data is subjected to facet detection to determine the facet relationship between multiple screw parts and the mounting structure surface;

[0096] The disassembly monitoring data is used to perform positioning detection to determine the planar positions of multiple screw parts on the corresponding mounting structure surface;

[0097] The disassembly monitoring data is used to identify the order of installation and determine the installation order of the multiple installation surfaces.

[0098] By combining multiple facet relationships, multiple planar positions, and multiple structural installation sequences, facet positioning data for multiple screw parts is generated.

[0099] Furthermore, the machine vision-based industrial parts inspection method also includes the following steps:

[0100] Step S104: Generate and display multiple corresponding surface structure contours according to multiple installation structure surfaces, and place multiple screw parts in the multiple surface structure contours according to multiple facet positioning data.

[0101] In this embodiment of the invention, projection planning is performed on multiple installation structural surfaces to determine the projection normal directions of the multiple installation structural surfaces. The three-dimensional structural model is projected according to the multiple projection normal directions to generate surface structural outlines corresponding to the multiple installation structural surfaces. Based on the multiple structural installation orders in the multiple facet positioning data, the display positions of the multiple surface structural outlines are planned. Then, according to the multiple display positions, the multiple surface structural outlines are displayed separately on the receiving plate. Furthermore, according to the multiple facet relationships and multiple plane positions in the multiple facet positioning data, the facet placement positions corresponding to multiple screw parts are planned on the multiple surface structural outlines. Then, according to the multiple facet placement positions, the multiple screw parts are facet-positioned and placed so that each screw part can be placed at the plane position of the surface structural outline of the corresponding installation structural surface.

[0102] Specifically, in another preferred embodiment provided by the present invention, the step of generating and displaying multiple corresponding surface structure contours according to multiple installation structure surfaces, and positioning and placing multiple screw parts on the multiple surface structure contours according to multiple facet positioning data, specifically includes the following steps:

[0103] Project the three-dimensional structural model according to the multiple installation structural surfaces to generate surface structural contours corresponding to the multiple installation structural surfaces;

[0104] Based on the multiple facet positioning data, plan the display positions of the multiple facet structure contours;

[0105] According to the multiple display positions, the multiple surface structure outlines are distinguished and displayed on the receiving plate;

[0106] Based on the multiple facet positioning data, plan the corresponding facet placement positions on the multiple facet structure contours;

[0107] According to the multiple faceted placement positions, multiple screw parts are positioned and placed in faceted positions.

[0108] Furthermore, the machine vision-based industrial parts inspection method also includes the following steps:

[0109] Step S105: Perform visual inspection on the multiple screw parts that are positioned and placed in a faceted manner to determine whether there are any visual abnormalities, and replace them accordingly if there are any visual abnormalities.

[0110] In this embodiment of the invention, after the faceted positioning and placement of multiple screw parts are completed, the multiple screw parts are inspected and photographed to obtain inspection and photographing data. The inspection and photographing data is then used for model identification to determine the screw model corresponding to the multiple screw parts. Based on the multiple screw models, standard images corresponding to the multiple screw parts are matched from a preset standard database. Based on the multiple standard images, the inspection and photographing data are compared and inspected to determine whether there is an appearance abnormality. If an appearance abnormality is determined, the abnormal screw with the appearance abnormality is selected from the multiple screw parts, and a corresponding replacement screw is matched according to the screw model corresponding to the abnormal screw. Replacement placement control is performed to remove the abnormal screw and replace it with a replacement screw, thereby realizing the automatic replacement of the abnormal screw.

[0111] It is understandable that screw parts may have abnormal appearances, including dents, scratches, cracks, rust, or bending.

[0112] Specifically, in another preferred embodiment provided by the present invention, the step of performing visual inspection on multiple screw parts arranged in a faceted positioning manner to determine whether there are any visual abnormalities, and replacing them accordingly when there are visual abnormalities, specifically includes the following steps:

[0113] Multiple screw parts with faceted positioning are inspected and photographed to obtain inspection and photographic data;

[0114] Obtain standard images of multiple screw parts;

[0115] Based on multiple standard images, the detection and capture data are compared and detected to determine whether there are any appearance abnormalities.

[0116] When there is an appearance abnormality, select the abnormal screw from the plurality of screw parts;

[0117] Match the replacement screw to the abnormal screw;

[0118] Perform replacement placement control, remove the abnormal screw, and replace it with the replacement screw.

[0119] Furthermore, Figure 2 The diagram illustrates the application architecture of the machine vision-based industrial parts inspection system provided in this embodiment of the invention.

[0120] Specifically, in another preferred embodiment of the present invention, an industrial parts inspection system based on machine vision includes:

[0121] The structure installation recognition module 101 is used to perform initial shooting, identify the target industrial product, match the three-dimensional structural model of the target industrial product, and perform structure installation recognition, dividing multiple installation structure surfaces.

[0122] In this embodiment of the invention, before performing manual disassembly and assembly operations for product repair, the product to be disassembled and assembled needs to be placed on the disassembly and assembly workbench. The structure installation recognition module 101 initializes the disassembly and assembly workbench by taking an initial photograph to obtain initial photograph data. Based on preset product feature data, it performs feature recognition and comparative analysis on the initial photograph data to determine the target industrial product. Then, it matches the three-dimensional structure model of the target industrial product in the preset model database. By automatically rotating the three-dimensional structure model at multiple angles, it performs planar recognition on the three-dimensional structure model at different rotation angles to determine multiple model structure planes on the surface of the three-dimensional structure model. Then, it performs structure installation recognition on the multiple model structure planes and selects multiple installation structure surfaces with screw assembly from the multiple model structure planes.

[0123] Specifically, in another preferred embodiment provided by the present invention, the structure installation identification module 101 specifically includes:

[0124] Initialize the shooting unit to perform initial shooting and acquire initial shooting data;

[0125] A target recognition unit is used to perform target recognition on the initial captured data to determine the target industrial product;

[0126] The model matching unit is used to match the three-dimensional structural model of the target industrial product in a preset model database.

[0127] A plane recognition unit is used to perform plane recognition on the three-dimensional structural model and determine multiple model structural planes;

[0128] The structural surface selection unit is used to identify the structural installation of multiple model structural planes and select multiple installation structural planes with screw assembly from the multiple model structural planes.

[0129] Furthermore, the machine vision-based industrial parts inspection system also includes:

[0130] The real-time monitoring and processing module 102 is used to monitor the manual disassembly process of the target industrial product in real time, acquire disassembly monitoring data, and collect screws and parts during the disassembly process.

[0131] In this embodiment of the invention, when workers are performing manual disassembly and assembly operations on the target industrial product for product repair, the real-time monitoring and processing module 102 monitors the manual disassembly process of the target industrial product in real time, acquires disassembly monitoring data, and generates a receiving and collection instruction. According to the receiving and collection instruction, the receiving plate is opened, and the screws and parts in the disassembly process of the target industrial product are received and collected through the receiving plate.

[0132] The facet positioning detection module 103 is used to perform facet positioning detection on the disassembly monitoring data based on multiple installation structure faces, and to obtain facet positioning data of multiple screw parts.

[0133] In this embodiment of the invention, the facet positioning detection module 103 performs facet detection on the disassembly monitoring data to determine the facet relationships between multiple screw parts and multiple mounting structure surfaces, and performs positioning detection on the disassembly monitoring data to determine the planar positions of multiple screw parts on the corresponding mounting structure surfaces. At the same time, it performs sequence identification on the disassembly monitoring data to determine the structural disassembly sequence corresponding to multiple mounting structure surfaces during the disassembly of the target industrial product, and reverses the structural disassembly sequence corresponding to multiple mounting structure surfaces to obtain the structural installation sequence corresponding to multiple mounting structure surfaces. Then, it comprehensively organizes the multiple facet relationships, multiple planar positions, and multiple structural installation sequences to generate facet positioning data for multiple screw parts.

[0134] Specifically, in another preferred embodiment provided by the present invention, the facet positioning and detection module 103 specifically includes:

[0135] The facet detection unit is used to perform facet detection on the disassembly monitoring data to determine the facet relationship between multiple screw parts and the mounting structure surface.

[0136] The positioning detection unit is used to perform positioning detection on the disassembly monitoring data to determine the planar position of multiple screw parts on the corresponding mounting structure surface;

[0137] The order identification unit is used to identify the order of the disassembly monitoring data and determine the structural installation order corresponding to the multiple installation structural surfaces.

[0138] The facet positioning data generation unit is used to integrate multiple facet relationships, multiple planar positions, and multiple structural installation sequences to generate facet positioning data for multiple screw parts.

[0139] Furthermore, the machine vision-based industrial parts inspection system also includes:

[0140] The faceted positioning and placement module 104 is used to generate and display multiple corresponding faceted structure contours according to multiple installation structure surfaces, and to perform faceted positioning and placement of multiple screw parts on the multiple faceted structure contours according to multiple faceted positioning data.

[0141] In this embodiment of the invention, the faceted positioning and placement module 104 projects and plans multiple installation structural surfaces, determines the projection normal directions of the multiple installation structural surfaces, projects the three-dimensional structural model according to the multiple projection normal directions, generates the surface structural outlines corresponding to the multiple installation structural surfaces, and plans the display positions of the multiple surface structural outlines according to the multiple structural installation order in the multiple faceted positioning data. Then, according to the multiple display positions, the multiple surface structural outlines are displayed separately on the receiving plate. Furthermore, according to the multiple faceted relationships and multiple plane positions in the multiple faceted positioning data, the faceted placement positions corresponding to multiple screw parts are planned on the multiple surface structural outlines. Then, according to the multiple faceted placement positions, the multiple screw parts are faceted and positioned, so that each screw part can be placed at the plane position of the surface structural outline of the corresponding installation structural surface.

[0142] The abnormal replacement placement module 105 is used to perform visual inspection on multiple screw parts that are positioned faceted, determine whether there are any visual abnormalities, and replace them accordingly when there are visual abnormalities.

[0143] In this embodiment of the invention, after the faceted positioning and placement of multiple screw parts are completed, the abnormal replacement placement module 105 detects and photographs the multiple screw parts, acquires detection and photographing data, and performs model identification on the detection and photographing data to determine the screw model corresponding to the multiple screw parts. Based on the multiple screw models, it matches the standard images corresponding to the multiple screw parts from a preset standard database. Based on the multiple standard images, it compares and detects the detection and photographing data to determine whether there is an appearance abnormality. If an appearance abnormality is determined, it selects the abnormal screw with the appearance abnormality from the multiple screw parts, matches the corresponding replacement screw according to the screw model of the abnormal screw, performs replacement placement control, removes the abnormal screw, and replaces and places the replacement screw, thereby realizing the automatic replacement of the abnormal screw.

[0144] Specifically, in another preferred embodiment provided by the present invention, the abnormal replacement placement module 105 specifically includes:

[0145] The detection and imaging unit is used to detect and photograph multiple screw parts that are positioned in a faceted manner, and to acquire detection and imaging data.

[0146] A standard image acquisition unit is used to acquire standard images of multiple screw parts;

[0147] The comparison detection unit is used to compare and detect the detection data based on multiple standard images to determine whether there are any appearance abnormalities.

[0148] An abnormality selection unit is used to select an abnormal screw from a plurality of screw parts when there is an appearance abnormality;

[0149] A replacement matching unit is used to match the replacement screw corresponding to the abnormal screw;

[0150] The replacement placement control unit is used to perform replacement placement control, remove the abnormal screw, and replace the replacement screw.

[0151] The embodiments described above are merely illustrative of several implementations of the present invention, and while the descriptions are specific and detailed, they should not be construed as limiting the scope of the present invention. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of the present invention, and these modifications and improvements all fall within the scope of protection of the present invention. Therefore, the scope of protection of this patent should be determined by the appended claims.

Claims

1. A machine vision-based method for inspecting industrial parts, characterized in that, The method specifically includes the following steps: Initialize the shooting, identify the target industrial product, match the three-dimensional structural model of the target industrial product, identify the structural installation, and divide multiple installation structural surfaces; The manual disassembly process of the target industrial product is monitored in real time to obtain disassembly monitoring data, and screws and parts collected during the disassembly process are collected. Based on multiple installation structure surfaces, the disassembly monitoring data is subjected to surface positioning detection to obtain surface positioning data of multiple screw parts; Based on multiple installation structural surfaces, multiple corresponding surface structure contours are generated and displayed. Based on multiple facet positioning data, multiple screw parts are facet positioning and placement on the multiple surface structure contours. Perform visual inspection on multiple screw parts that are positioned and placed in a faceted manner to determine if there are any visual abnormalities, and replace them accordingly if there are any. The initial shooting, identification of the target industrial product, matching of the target industrial product's three-dimensional structural model, and identification of structural installation, including the division of multiple installation structural surfaces, specifically includes the following steps: Perform initial shooting and acquire initial shooting data; Target identification is performed on the initial captured data to determine the target industrial product; In a pre-set model database, match the three-dimensional structural model of the target industrial product. Perform plane identification on the three-dimensional structural model to determine multiple model structural planes; Structural installation identification is performed on multiple model structural planes, and multiple installation structural planes with screw assembly are selected from the multiple model structural planes.

2. The machine vision-based industrial parts inspection method according to claim 1, characterized in that, The process of manually disassembling the target industrial product in real time, acquiring disassembly monitoring data, and collecting screws and other parts during the disassembly process specifically includes the following steps: Real-time monitoring of the manual disassembly process of target industrial products to obtain disassembly monitoring data; Generate acceptance set instructions; According to the receiving and collection instructions, open the receiving plate and use the receiving plate to receive and collect the screws and parts during the disassembly process.

3. The machine vision-based industrial parts inspection method according to claim 1, characterized in that, The step of performing facet positioning detection on the disassembly monitoring data based on multiple installation structure surfaces to obtain facet positioning data of multiple screw parts specifically includes the following steps: The disassembly monitoring data is subjected to facet detection to determine the facet relationship between multiple screw parts and the mounting structure surface; The disassembly monitoring data is used to perform positioning detection to determine the planar positions of multiple screw parts on the corresponding mounting structure surface; The disassembly monitoring data is used to identify the order of installation and determine the installation order of the multiple installation structural surfaces. By combining multiple facet relationships, multiple planar positions, and multiple structural installation sequences, facet positioning data for multiple screw parts is generated.

4. The machine vision-based industrial parts inspection method according to claim 2, characterized in that, The process of generating and displaying multiple corresponding surface structure contours based on multiple installation structural surfaces, and positioning and placing multiple screw parts on the multiple surface structure contours according to multiple facet positioning data, specifically includes the following steps: Project the three-dimensional structural model according to the multiple installation structural surfaces to generate surface structural contours corresponding to the multiple installation structural surfaces; Based on the multiple facet positioning data, plan the display positions of the multiple facet structure contours; According to the multiple display positions, the multiple surface structure outlines are distinguished and displayed on the receiving plate; Based on the multiple facet positioning data, plan the corresponding facet placement positions on the multiple facet structure contours; According to the multiple faceted placement positions, multiple screw parts are positioned and placed in faceted positions.

5. The machine vision-based industrial parts inspection method according to claim 1, characterized in that, The process of visually inspecting multiple screw parts positioned on facets to determine if there are any abnormalities, and then replacing them accordingly if abnormalities are found, specifically includes the following steps: Multiple screw parts with faceted positioning are inspected and photographed to obtain inspection and photographic data; Obtain standard images of multiple screw parts; Based on multiple standard images, the detection and capture data are compared and detected to determine whether there are any appearance abnormalities. When there is an appearance abnormality, select the abnormal screw from the plurality of screw parts; Match the replacement screw to the abnormal screw; Perform replacement placement control, remove the abnormal screw, and replace it with the replacement screw.

6. A machine vision-based industrial parts inspection system, characterized in that, The system includes a structure installation identification module, a real-time monitoring and processing module, a faceted positioning detection module, a faceted positioning placement module, and an abnormal placement replacement module, wherein: The structural installation recognition module is used to perform initial shooting, identify the target industrial product, match the three-dimensional structural model of the target industrial product, and perform structural installation recognition, dividing multiple installation structural surfaces. The real-time monitoring and processing module is used to monitor the manual disassembly process of the target industrial product in real time, acquire disassembly monitoring data, and collect screws and parts during the disassembly process. The facet positioning detection module is used to perform facet positioning detection on the disassembly monitoring data based on multiple installation structure faces, and to obtain facet positioning data of multiple screw parts. The faceted positioning and placement module is used to generate and display multiple corresponding faceted structure contours according to multiple installation structure surfaces, and to perform faceted positioning and placement of multiple screw parts on the multiple faceted structure contours according to multiple faceted positioning data. The abnormal replacement placement module is used to perform visual inspection on multiple screw parts that are placed in a faceted positioning manner, determine whether there are any visual abnormalities, and replace them accordingly when there are visual abnormalities. The structure installation identification module specifically includes: Initialize the shooting unit to perform initial shooting and acquire initial shooting data; A target recognition unit is used to perform target recognition on the initial captured data to determine the target industrial product; The model matching unit is used to match the three-dimensional structural model of the target industrial product in a preset model database. A plane recognition unit is used to perform plane recognition on the three-dimensional structural model and determine multiple model structural planes; The structural surface selection unit is used to identify the structural installation of multiple model structural planes and select multiple installation structural planes with screw assembly from the multiple model structural planes.

7. The machine vision-based industrial parts inspection system according to claim 6, characterized in that, The faceted positioning and detection module specifically includes: The facet detection unit is used to perform facet detection on the disassembly monitoring data to determine the facet relationship between multiple screw parts and the mounting structure surface. The positioning detection unit is used to perform positioning detection on the disassembly monitoring data to determine the planar position of multiple screw parts on the corresponding mounting structure surface; The order recognition unit is used to identify the order of the disassembly monitoring data and determine the structural installation order corresponding to the multiple installation structural surfaces. The facet positioning data generation unit is used to integrate multiple facet relationships, multiple planar positions, and multiple structural installation sequences to generate facet positioning data for multiple screw parts.

8. The machine vision-based industrial parts inspection system according to claim 6, characterized in that, The abnormal replacement placement module specifically includes: The detection and imaging unit is used to detect and photograph multiple screw parts that are positioned in a faceted manner, and to acquire detection and imaging data. A standard image acquisition unit is used to acquire standard images of multiple screw parts; The comparison detection unit is used to compare and detect the detection data based on multiple standard images to determine whether there are any appearance abnormalities. An anomaly selection unit is used to select an abnormal screw from a plurality of screw parts when there is an appearance anomaly. A replacement matching unit is used to match the replacement screw corresponding to the abnormal screw; The replacement placement control unit is used to perform replacement placement control, remove the abnormal screw, and replace the replacement screw.

Images