A visual inspection system and method for detecting mis-installed or missing corner pieces in aircraft assembly.

Abstract

This invention discloses a visual inspection system and method for detecting misinstalled or missing corner pieces during aircraft assembly. The visual inspection device comprises an attitude adjustment module and a vision module carried by a mobile carrier, which move between various inspection stations. The attitude adjustment module adjusts the shooting angle of the vision module through three-axis movement. The vision module captures images of the aircraft components. A central control system controls the attitude adjustment module and the vision module, enabling manual selection, position correction, contour recognition, contour comparison, and detection database recording of the aircraft component images through the central control system's display interface. In calibration mode, the central control system acquires inspection templates for the aircraft components with installed corner pieces at each inspection station, forming an inspection template database. In inspection mode, the images of the aircraft components to be inspected captured at each inspection station are matched with the corresponding inspection templates to obtain the corner piece inspection results for each inspection station.

Description

Technical Field

[0001] This invention relates to, but is not limited to, the field of visual inspection technology, and in particular to a visual inspection system and method for detecting mis-installed or missing corner pieces during aircraft assembly. Background Technology

[0002] Aircraft components such as wings and fuselages contain numerous and varied parts, including brackets, used to secure piping, wiring harnesses, or finished components. The correct installation of these brackets directly affects the smooth installation of the piping system; omissions, incorrect installation positions, or incorrect installation directions can significantly impact aircraft flight safety.

[0003] Due to the small size, large number, and irregularity of corner pieces, the installation position, orientation, and quantity of aircraft components are currently mainly checked manually after manufacturing. Traditional manual inspection methods are inefficient, the results depend on the worker's condition and skill level, and prolonged work can lead to worker fatigue. Furthermore, manual inspection cannot quantitatively define the extent of misassembly, making it difficult to effectively detect these errors or omissions. Summary of the Invention

[0004] The purpose of this invention is to address the aforementioned technical problems by providing a visual inspection system and method for detecting mis-installed corner pieces in aircraft assembly. This addresses the shortcomings of existing inspection methods for corner piece-type parts in aircraft assembly, such as low assembly efficiency, reliance on worker condition and skill level for inspection results, and the risk of worker fatigue due to prolonged work. Furthermore, manual inspection cannot quantitatively define the mis-installation status of parts, making it difficult to effectively detect these mis-installation issues.

[0005] The technical solution of the present invention: In a first aspect, the present invention provides a visual inspection system for mis-installed corner pieces in aircraft assembly, comprising: a visual inspection device and a control system; wherein, the visual inspection device comprises: a mobile carrier, an attitude adjustment module fixedly installed on the mobile carrier, and a visual module installed at the end of the attitude adjustment module;

[0006] The mobile carrier is used to carry the attitude adjustment module and the vision module and move between each detection station; the attitude adjustment module is used to adjust the shooting angle of the vision module by moving the vertical axis, pitch axis and rotation axis; the vision module is used to capture images of the machine parts and has focusing and multi-angle lighting functions.

[0007] The overall control system is connected to the posture adjustment module and the vision module in the vision inspection device, respectively, and is used to control the posture adjustment module and the vision module. The overall control system displays images of the captured body parts and performs functions such as manual selection, position correction, contour recognition, contour comparison and detection database recording.

[0008] The overall control system has a calibration mode or a detection mode. In calibration mode, it works with a vision inspection device to acquire the detection template for each detection station for a machine body component with correctly installed corner pieces, and forms a detection template database by combining the layout of all detection stations. In detection mode, it works with a vision inspection device to match the images of the machine body component to be inspected taken at each detection station with the detection templates at the corresponding detection stations to obtain the corner piece detection results for each detection station.

[0009] Optionally, in the visual inspection system for mis-assembly of corner pieces in aircraft assembly as described above, the central control system has the functions of calling the inspection template database, controlling the attitude adjustment module to adjust the attitude, controlling the vision module to capture images and complete the image-based mis-assembly identification and analysis, and storing and analyzing the inspection results.

[0010] Secondly, embodiments of the present invention also provide a visual inspection method for mis-installed or missing corner pieces during aircraft assembly, wherein the visual inspection method is performed using a visual inspection system as described in any of the above claims, and the method includes:

[0011] Step 1: Start the visual inspection device and enter calibration mode;

[0012] Step 2: Secure the machine body parts with the corner plates correctly installed to the pre-planned placement position using the transfer device, and pull the vision inspection device to the pre-planned inspection position;

[0013] Step 3: Control the attitude of each axis in the attitude adjustment module of the vision inspection device through the overall control system, and adjust the shooting angle according to the real-time image fed back by the vision module; after determining the shooting angle at the current inspection station, take pictures of the machine parts and generate machine part images; display the machine part images on the display interface of the overall control system, and select the position correction area in the machine part images to correct the image movement, rotation and scaling interference caused by the positioning deviation of the vision module in subsequent inspections;

[0014] Step 4: Select the corner piece area from the machine part images displayed on the main control system display interface, match the drawing number of the corner piece area with the corner piece number, and automatically identify and record the corner piece outline;

[0015] Step 5: Record the attitude information of the attitude adjustment module, the captured images of the machine parts, the position and size information of the position correction area block diagram, the structural outline information within the position correction area block diagram, the position and size information of the corner piece area block diagram, the structural outline information within the corner piece area block diagram, and the number information of the corresponding corner piece in the corner piece area block diagram, to obtain the detection template at the current detection station.

[0016] Step six: Move the visual inspection device to the next pre-planned inspection station;

[0017] Step 7: Repeat steps 3 to 6 until a detection template for each detection station is obtained, forming a detection template database, and completing the calibration of the visual inspection device; the detection template database includes: the layout of detection stations for the machine parts to be inspected, and the detection template for each detection station;

[0018] Step 8: For the machine parts to be inspected, the vision inspection device is controlled by the central control system to enter the error and omission detection mode.

[0019] Step 9: Fix the machine part to be inspected to the pre-planned placement position using the transfer device, and pull the vision inspection device to the pre-planned inspection position;

[0020] Step 10: The detection template of the current detection station is called through the central control system, so that the visual inspection device automatically adjusts the posture information recorded in the detection template, automatically controls the vision module to capture images of the machine parts to be inspected, automatically corrects the position of the corner piece area diagram in the image of the machine parts to be inspected based on the position correction area diagram in the detection template, automatically identifies the corner piece outline and obtains the recognition result (OK / NG) of each corner piece area diagram by comparing it with the detection template, and then obtains the detection results of mis-installed or missing corner pieces in the image of the machine parts to be inspected, thus completing the visual inspection of the current detection station;

[0021] Step 11: Move the visual inspection device to the next inspection station and repeat the inspection process in Step 10.

[0022] Step 12: Repeat step 11 until the detection results of the corresponding detection station are obtained at each detection station, so as to complete the detection of misaligned or missing corner pieces of the entire machine body component under inspection and output the detection results.

[0023] Alternatively, in the method described above,

[0024] In step two, the relative positions between the pre-planned placement stations and each detection station are fixed. The planning of each detection station is obtained manually, and whether to place the detection station is determined based on the imaging effect of the vision module under each detection station.

[0025] Alternatively, in the method described above,

[0026] In step three, the method of adjusting the shooting angle based on the real-time image feedback from the vision module is as follows: the central control system controls each motion axis in the posture adjustment module based on the real-time captured images of the machine parts to achieve the adjustment of the shooting angle.

[0027] Optionally, in the method described above, in step three,

[0028] After the position correction area is selected in the image of the body component, the central control system automatically identifies the contour information of the body structure image within the position correction area. This information is then used for subsequent inspection of the body component to be inspected, where the position correction area at the current inspection station is matched with the image of the body component to be inspected.

[0029] The selection principle for the position correction area is to select structures with distinct geometric features, such as intersections and slide rails in the machine body components, which are less prone to assembly errors.

[0030] Alternatively, in the method described above,

[0031] The method for selecting the corner piece area in the machine part image in step four is as follows: select the corner piece to be inspected according to the detection requirements, and the shape, size and position of the corner piece selection box can be adjusted.

[0032] Optionally, the method described above for step six, moving the visual inspection device, is as follows: the visual inspection device is placed into the inspection positioning frame of the next inspection station by means of manual traction or automatic guidance by a mobile device, with a positioning accuracy at the centimeter level.

[0033] In step nine, the placement of the machine body parts to be inspected and the visual inspection device is as follows: the traction device manually or automatically positions them into the corresponding positioning frame, with a positioning accuracy on the order of centimeters.

[0034] Optionally, in the method described above, the central control system has the functions of calling the detection template database, controlling the posture adjustment module to adjust the posture, controlling the vision module to capture images and complete image-based error / omission identification and analysis, and storing the analysis results; the central control system performs image-based error / omission identification and analysis, including:

[0035] S1, call the detection template recorded at the current detection station;

[0036] S2, perform preliminary contour extraction on the images of the machine parts captured at the current detection station;

[0037] S3, by combining the structural contour within the position correction region block diagram in the detection template with the contour information of the captured image of the machine part to be inspected, the position correction region within the position correction region block diagram in the image of the machine part to be inspected is identified, and the displacement and rotation angle of the position correction region in the two images are calculated.

[0038] S4. Based on the calculated displacement and rotation angle of the position correction area, correct the position of the corner area block diagram in the detection template in the newly captured image of the body part to be inspected.

[0039] S5, identify corner piece outline information within each corner piece region bounding diagram;

[0040] S6. Match the identified corner piece contour information with the detection template. If the matching similarity exceeds the set threshold, the corner piece is judged to be installed normally and OK is output. If the matching similarity is less than the set threshold due to missing parts, incorrect corner piece model, or incorrect corner piece installation direction, the corner piece is judged to be installed incorrectly and NG is output.

[0041] S7 outputs and displays the analysis results of all corner pieces in the captured image of the machine body component to be inspected, completing the corner piece inspection at the current station position.

[0042] Alternatively, in the method described above,

[0043] The detection results in step 12 include detection station information, images of the machine body parts to be inspected, detection results (OK or NG) for each corner piece in the entire machine body parts to be inspected, and the number of the incorrectly installed corner piece.

[0044] The beneficial effects of this invention: This invention provides a visual inspection system and method for detecting misinstalled or missing corner pieces in aircraft assembly. Targeting the numerous and varied types of corner pieces present in aircraft assembly, it provides a visual inspection system comprising a visual inspection device and a central control system. The inspection device can be moved between inspection stations and includes an attitude adjustment module and a vision module. The central control system controls each module within the inspection device and has calibration and inspection modes. Utilizing a detection template database generated in the calibration mode, it automatically detects misinstalled or missing corner pieces in localized areas of the aircraft body component under inspection in the inspection mode. Through the movement of the visual inspection system between inspection stations, it automatically detects misinstalled or missing corner pieces in the entire aircraft body component under inspection and outputs information such as the incorrectly installed corner piece number. The visual inspection system and method proposed in this invention replace manual inspection of misinstalled or missing corner pieces, accurately and quickly assessing the integrity and accuracy of component assembly, and improving the digitalization level of aircraft assembly. Attached Figure Description

[0045] The accompanying drawings are provided to further understand the technical solutions of the present invention and constitute a part of the specification. They are used together with the embodiments of this application to explain the technical solutions of the present invention and do not constitute a limitation on the technical solutions of the present invention.

[0046] Figure 1 This is a schematic diagram of a wing box system corner piece misassembly detection provided by an embodiment of the present invention;

[0047] Figure 2 This is a schematic diagram of a visual inspection device in a visual inspection system for mis-installed or missing corner pieces in aircraft assembly, provided by an embodiment of the present invention.

[0048] Figure 3This is a schematic diagram of the position correction area and corner piece area of ​​the visual inspection system provided in the embodiment of the present invention in calibration mode for the selected area of ​​the captured body structure image;

[0049] Figure 4 This is a schematic diagram illustrating how the visual inspection system provided in this embodiment of the invention identifies the corner outline and forms a corner outline template for an image within a corner area bounding box in calibration mode.

[0050] Figure 5 This is a schematic diagram illustrating the process of determining the correct installation of corner pieces in the detection mode of the visual inspection system provided in this embodiment of the invention.

[0051] Figure 6 This is a schematic diagram illustrating the process of determining incorrect or missing corner pieces in the detection mode of the visual inspection system provided in this embodiment of the invention. Detailed Implementation

[0052] To make the objectives, technical solutions, and advantages of the present invention clearer, the embodiments of the present invention will be described in detail below with reference to the accompanying drawings. It should be noted that, unless otherwise specified, the embodiments and features described in this application can be arbitrarily combined with each other.

[0053] As explained in the background section, corner brackets and other components play a crucial role in aircraft assembly, exhibiting unique characteristics. Existing manual inspection methods for correct corner bracket installation also present numerous problems, including low inspection efficiency, reliance on worker condition and skill level for results, worker fatigue from prolonged work, and the inability to quantitatively define misinstallation, making it difficult to effectively detect errors or omissions. These issues can lead to defective components flowing into subsequent processes, affecting the installation of subsequent piping, wiring harnesses, and finished products.

[0054] Based on the aforementioned problems with manual inspection methods, and considering the rapid development of machine vision inspection technology in recent years as a new type of digital inspection technology with advantages such as non-contact operation, high accuracy, and easy data storage and traceability, this invention aims to improve the accuracy of system structural component inspection by applying digital inspection methods. Therefore, this invention provides a visual inspection method for mis-installed or missing corner pieces in aircraft assembly. By introducing machine vision inspection technology into the digital inspection of aerospace components, it replaces manual inspection of complex components, accurately and quickly assesses the integrity and accuracy of component assembly, and further enhances the digitalization level of aircraft assembly.

[0055] To address the aforementioned problems and the bearing end reduction requirements for elongated parts, embodiments of the present invention provide a constant force compression device and method for achieving bearing end reduction in elongated parts.

[0056] To address the aforementioned problems and the bearing end reduction requirements for elongated parts, embodiments of the present invention provide a constant force compression device and method for achieving bearing end reduction in elongated parts.

[0057] In order to accurately and quickly assess the integrity and accuracy of corner piece assembly, which is a large number of corner pieces and other parts in aircraft assembly, this invention proposes a visual inspection system and method for detecting mis-assembly and omission of corner pieces in aircraft assembly, realizing automatic detection of mis-assembly and omission of corner pieces in the system.

[0058] This invention provides a visual inspection system for detecting mis-installed or missing corner pieces during aircraft assembly. The visual inspection system includes a visual inspection device and a central control system. Figure 2 This is a schematic diagram of a visual inspection device in a visual inspection system for mis-installed or missing corner pieces during aircraft assembly, provided by an embodiment of the present invention. Figure 3 As shown, the visual detection device in this embodiment of the invention includes: a mobile carrier, an attitude adjustment module fixedly installed on the mobile carrier, and a visual module installed at the end of the attitude adjustment module.

[0059] like Figure 2 In the structure of the visual inspection device shown, the mobile carrier is used to carry the attitude adjustment module and the vision module and move between each inspection station; the attitude adjustment module is used to adjust the shooting angle of the vision module by moving the vertical axis, pitch axis and rotation axis; the vision module is used to take images of the machine parts and has focusing and multi-angle lighting functions.

[0060] In this embodiment of the invention, the overall control system is connected to the posture adjustment module and the vision module in the vision inspection device, respectively, and is used to control the posture adjustment module and the vision module. The overall control system displays images of the captured machine parts and performs functions such as manual selection, position correction, contour recognition, contour comparison and detection database recording.

[0061] In addition, the overall control system provided in this embodiment of the invention has a calibration mode or a detection mode:

[0062] On the one hand, in calibration mode, in conjunction with a visual inspection device, the body parts with correctly installed corner pieces are used to obtain the inspection template of each inspection station at each inspection station, and combined with the inspection template database of all inspection stations;

[0063] On the other hand, in the detection mode, in conjunction with the visual inspection device, the images of the machine parts to be inspected, captured at each inspection station, are matched with the inspection template at the corresponding inspection station to obtain the corner piece detection results at each inspection station.

[0064] In one implementation of this invention, the overall control system of the visual inspection system has the functions of calling the inspection template database, controlling the posture adjustment module to adjust the posture, controlling the visual module to take pictures and complete the image-based error and omission identification and analysis, and storing and analyzing the inspection results.

[0065] Based on the visual inspection system for mis-installed corner pieces in aircraft assembly provided in the above embodiments of the present invention, the present invention also provides a visual inspection method for mis-installed corner pieces in aircraft assembly. The visual inspection method, performed using any of the visual inspection systems provided in the above embodiments, includes the following steps:

[0066] Step 1: Start the visual inspection device and enter calibration mode.

[0067] Step two, as Figure 1 As shown, the aircraft parts (wings, fuselage, etc.) with the corner plates correctly installed are fixed in the pre-planned placement position by the transfer device, and the vision inspection device is pulled to the pre-planned inspection position. Figure 1 This is a schematic diagram of a wing box system corner piece misassembly detection method provided in an embodiment of the present invention.

[0068] Step 3: Control the attitude of each axis in the attitude adjustment module of the vision inspection device through the central control system, and adjust the shooting angle according to the real-time images fed back by the vision module; after determining the shooting angle at the current inspection station position, photograph the machine body parts and generate... Figure 3 The image shows the machine body components; the machine body components are displayed on the interface of the central control system, and a selection is made within the machine body components image. Figure 3 The position correction region diagram shown is used to correct image movement, rotation, and scaling interference caused by visual module positioning deviations during subsequent detection. Figure 3 This is a schematic diagram of the position correction area and corner area of ​​the visual inspection system provided in the embodiment of the present invention in calibration mode for the selected area of ​​the captured body structure image.

[0069] Step 4: Select the machine parts from the images displayed on the main control system's display interface. Figure 3 The diagram shows the corner piece area, and the diagram number is matched with the corner piece number. The corner piece outline is automatically identified and recorded. The identification process is as follows: Figure 4 As shown, Figure 4 This is a schematic diagram illustrating how the visual inspection system provided in this embodiment of the invention identifies the corner outline and forms a corner outline template for an image within a corner area bounding box in calibration mode.

[0070] Step 5: Record the attitude information of the attitude adjustment module, the captured images of the machine parts, the position and size information of the position correction area block diagram, the structural outline information within the position correction area block diagram, the position and size information of the corner piece area block diagram, the structural outline information within the corner piece area block diagram, and the number information of the corresponding corner piece in the corner piece area block diagram, to obtain the detection template at the current detection station.

[0071] Step six: Move the visual inspection device to the next pre-planned inspection station;

[0072] Step 7: Repeat steps 3 to 6 until a detection template for each detection station is obtained, forming a detection template database, and completing the calibration of the visual inspection device; the detection template database includes: the layout of detection stations for the machine parts to be inspected, and the detection template for each detection station;

[0073] Step 8: For the machine parts to be inspected, the vision inspection device is controlled by the central control system to enter the error and omission detection mode.

[0074] Step nine, as Figure 1 As shown, the machine parts to be inspected are fixed in the pre-planned placement position by the transfer device, and the vision inspection device is pulled to the pre-planned inspection position.

[0075] Step 10: The system calls the detection template of the current detection station through the central control system, so that the vision inspection device automatically adjusts the posture information recorded in the detection template, automatically controls the vision module to capture images of the machine parts to be inspected, automatically corrects the position of the corner piece area diagram in the image of the machine parts to be inspected based on the position correction area diagram in the detection template, automatically identifies the corner piece outline and obtains the recognition result (OK / NG) of each corner piece area diagram by comparing it with the detection template.

[0076] Figure 5 This is a schematic diagram illustrating the process of determining the correct installation of corner pieces in the detection mode of the visual inspection system provided in this embodiment of the invention. Figure 6 This is a schematic diagram illustrating the process of determining incorrect or missing corner pieces in the detection mode of the visual inspection system provided in this embodiment of the invention.

[0077] like Figure 5 As shown, when the corner piece is installed correctly, the corner piece outline automatically identified within the corner piece frame is matched with the template for correlation. If the correlation is higher than the set threshold, the system determines that the corner piece is installed correctly and outputs "OK".

[0078] like Figure 6As shown, when problems such as incorrect corner piece model, correct corner piece model but incorrect installation direction, or missing corner piece are encountered, the corner piece outline identified within the corner piece frame becomes unrelated to the corner piece outline in the corresponding template. If the correlation is lower than the set threshold, the system determines that the corner piece is incorrectly installed and outputs NG.

[0079] At the current detection station, after performing error / omission detection on the bounding box of each corner piece in the body structure image, the error / omission detection results of each corner piece in the image are obtained, thus realizing error / omission detection at the current detection station.

[0080] Step 11: Move the visual inspection device to the next inspection station and repeat the inspection process in Step 10.

[0081] Step 12: Repeat step 11 until the detection results of the corresponding detection station are obtained at each detection station, so as to complete the detection of misaligned or missing corner pieces of the entire machine body component under inspection and output the detection results.

[0082] This completes the automatic detection of misinstallation or omission of all corner pieces within the machine body components to be inspected.

[0083] In one implementation of this invention, the relative positions between the pre-planned placement sites and each detection site in step two are fixed. The planning of each detection site is obtained manually, and whether to place the detection site is determined based on the imaging effect of the vision module under each detection site.

[0084] In one implementation of this invention, the method of adjusting the shooting angle based on the real-time image fed back by the vision module in step three is as follows: the central control system controls each motion axis in the posture adjustment module based on the real-time captured images of the machine parts to achieve the adjustment of the shooting angle.

[0085] In one implementation of this invention, in step three above, after selecting the position correction region frame in the machine component image, the central control system automatically identifies the contour information of the machine structure image within the position correction region frame, which is used for subsequent inspection of the machine component to be inspected, by matching the position correction region frame at the current inspection station with the image of the machine component to be inspected.

[0086] It should be noted that the selection principle for the position correction area is to select structures with distinct geometric features, such as intersections and slide rails in the machine body components, which are less prone to assembly errors.

[0087] In one implementation of this invention, the method for selecting the corner piece area in the body component image in step four is as follows: select the corner piece to be inspected according to the detection requirements, and the shape, size and position of the corner piece selection box are adjustable.

[0088] In one implementation of this invention, step six involves moving the visual inspection device by manually traction or by automatically guiding it into position using a mobile device, placing the visual inspection device into the detection positioning frame of the next inspection station with a positioning accuracy at the centimeter level.

[0089] Correspondingly, in step nine, the placement of the machine parts to be inspected and the visual inspection device is as follows: the traction device manually or automatically places them into the corresponding positioning frame, with a positioning accuracy on the order of centimeters.

[0090] In one implementation of this invention, the central control system has the functions of calling the detection template database, controlling the posture adjustment module to adjust the posture, controlling the vision module to capture images and complete image-based error / omission identification and analysis, and storing the analysis results; the central control system performs image-based error / omission identification and analysis, including:

[0091] S1, call the detection template recorded at the current detection station;

[0092] S2, perform preliminary contour extraction on the images of the machine parts captured at the current detection station;

[0093] S3, by combining the structural contour within the position correction region block diagram in the detection template with the contour information of the captured image of the machine part to be inspected, the position correction region within the position correction region block diagram in the image of the machine part to be inspected is identified, and the displacement and rotation angle of the position correction region in the two images are calculated.

[0094] S4. Based on the calculated displacement and rotation angle of the position correction area, correct the position of the corner area block diagram in the detection template in the newly captured image of the body part to be inspected.

[0095] S5, identify corner piece outline information within each corner piece region bounding diagram;

[0096] S6. Match the identified corner piece contour information with the detection template. If the matching similarity exceeds the set threshold, the corner piece is judged to be installed normally and OK is output. If the matching similarity is less than the set threshold due to missing parts, incorrect corner piece model, or incorrect corner piece installation direction, the corner piece is judged to be installed incorrectly and NG is output.

[0097] S7 outputs and displays the analysis results of all corner pieces in the captured image of the machine body component to be inspected, completing the corner piece inspection at the current station position.

[0098] In one implementation of this invention, the detection results in step twelve include detection station information, images of the machine body component to be inspected, detection results (OK or NG) for each corner piece in the entire machine body component to be inspected, and the number of the incorrectly installed corner piece.

[0099] While the embodiments disclosed in this invention are as described above, they are merely illustrative of the embodiments to facilitate understanding of the invention and are not intended to limit the invention. Any person skilled in the art to which this invention pertains may make any modifications and variations in the form and details of the implementation without departing from the spirit and scope disclosed herein; however, the scope of patent protection for this invention shall still be determined by the scope defined in the appended claims.

Claims

1. A visual inspection method for mis-assembly or omission of corner pieces in aircraft assembly, characterized in that, A visual inspection system is used to perform a visual inspection method. The visual inspection system includes a visual inspection device and a central control system. The method includes: Step 1: Start the visual inspection device and enter calibration mode; Step 2: Secure the machine body parts with the corner plates correctly installed to the pre-planned placement position using the transfer device, and pull the vision inspection device to the pre-planned inspection position; Step 3: Control the attitude of each axis in the attitude adjustment module of the vision inspection device through the overall control system, and adjust the shooting angle according to the real-time image fed back by the vision module; after determining the shooting angle at the current inspection station, take pictures of the machine parts and generate machine part images; display the machine part images on the display interface of the overall control system, and select the position correction area in the machine part images to correct the image movement, rotation and scaling interference caused by the positioning deviation of the vision module in subsequent inspections; Step 4: Select the corner piece area from the machine part images displayed on the main control system display interface, match the drawing number of the corner piece area with the corner piece number, and automatically identify and record the corner piece outline; Step 5: Record the attitude information of the attitude adjustment module, the captured images of the machine parts, the position and size information of the position correction area block diagram, the structural outline information within the position correction area block diagram, the position and size information of the corner piece area block diagram, the structural outline information within the corner piece area block diagram, and the number information of the corresponding corner piece in the corner piece area block diagram, to obtain the detection template at the current detection station. Step six: Move the visual inspection device to the next pre-planned inspection station; Step 7: Repeat steps 3 to 6 until a detection template for each detection station is obtained, forming a detection template database, and completing the calibration of the visual inspection device; the detection template database includes: the layout of detection stations for the machine parts to be inspected, and the detection template for each detection station; Step 8: For the machine parts to be inspected, the vision inspection device is controlled by the central control system to enter the error and omission detection mode. Step 9: Fix the machine part to be inspected to the pre-planned placement position using the transfer device, and pull the vision inspection device to the pre-planned inspection position; Step 10: The detection template of the current detection station is called through the central control system, so that the visual inspection device automatically adjusts the posture information recorded in the detection template, automatically controls the vision module to capture images of the machine parts to be inspected, automatically corrects the position of the corner piece area diagram in the image of the machine parts to be inspected based on the position correction area diagram in the detection template, automatically identifies the corner piece outline and obtains the recognition result (OK / NG) of each corner piece area diagram by comparing it with the detection template, and then obtains the detection results of mis-installed or missing corner pieces in the image of the machine parts to be inspected, thus completing the visual inspection of the current detection station; Step 11: Move the visual inspection device to the next inspection station and repeat the inspection process in Step 10. Step 12: Repeat step 11 until the detection results of the corresponding detection station are obtained at each detection station, so as to complete the detection of misaligned or missing corner pieces of the entire machine body component under inspection and output the detection results.

2. The method according to claim 1, characterized in that, In step two, the relative positions between the pre-planned placement stations and each detection station are fixed. The planning of each detection station is obtained manually, and whether to place the detection station is determined based on the imaging effect of the vision module under each detection station.

3. The method according to claim 1, characterized in that, In step three, the method of adjusting the shooting angle based on the real-time image feedback from the vision module is as follows: the central control system controls each motion axis in the posture adjustment module based on the real-time captured images of the machine parts to achieve the adjustment of the shooting angle.

4. The method according to claim 1, characterized in that, In step three After the position correction area is selected in the image of the body component, the central control system automatically identifies the contour information of the body structure image within the position correction area. This information is then used for subsequent inspection of the body component to be inspected, where the position correction area at the current inspection station is matched with the image of the body component to be inspected. The selection principle for the position correction area is to select structures with distinct geometric features, such as intersections and slide rails in the machine body components, which are less prone to assembly errors.

5. The method according to claim 1, characterized in that, The method for selecting the corner piece area in the machine part image in step four is as follows: select the corner piece to be inspected according to the detection requirements, and the shape, size and position of the corner piece selection box can be adjusted.

6. The method according to claim 1, characterized in that, The method for moving the visual inspection device in step six is ​​as follows: the visual inspection device is placed into the inspection positioning frame of the next inspection station by means of manual traction or automatic guidance of the mobile device, and its positioning accuracy is at the centimeter level. In step nine, the placement of the machine body parts to be inspected and the visual inspection device is as follows: the traction device manually or automatically positions them into the corresponding positioning frame, with a positioning accuracy on the order of centimeters.

7. The method according to claim 1, characterized in that, The overall control system has the functions of calling the detection template database, controlling the posture adjustment module to adjust the posture, controlling the vision module to capture images and complete the image-based error and omission identification and analysis, and storing the analysis results; The central control system performs image-based error / missing item identification and analysis, including: S1, call the detection template recorded at the current detection station; S2, perform preliminary contour extraction on the images of the machine parts captured at the current detection station; S3, by combining the structural contour within the position correction region block diagram in the detection template with the contour information of the captured image of the machine part to be inspected, the position correction region within the position correction region block diagram in the image of the machine part to be inspected is identified, and the displacement and rotation angle of the position correction region in the two images are calculated. S4. Based on the calculated displacement and rotation angle of the position correction area, correct the position of the corner area block diagram in the detection template in the newly captured image of the body part to be inspected. S5, identify corner piece outline information within each corner piece region bounding diagram; S6. Match the identified corner piece contour information with the detection template. If the matching similarity exceeds the set threshold, the corner piece is judged to be installed normally and OK is output. If the matching similarity is less than the set threshold due to missing parts, incorrect corner piece model, or incorrect corner piece installation direction, the corner piece is judged to be installed incorrectly and NG is output. S7 outputs and displays the analysis results of all corner pieces in the captured image of the machine body component to be inspected, completing the corner piece inspection at the current station position.

8. The method according to claim 1, characterized in that, The detection results in step 12 include detection station information, images of the machine body parts to be inspected, detection results of each corner piece in the entire machine body parts to be inspected, and the numbers of incorrectly installed corner pieces.

9. The method according to claim 1, characterized in that, The visual inspection device in the visual inspection system includes: a mobile carrier, an attitude adjustment module fixedly installed on the mobile carrier, and a visual module installed at the end of the attitude adjustment module. The mobile carrier is used to carry the attitude adjustment module and the vision module and move between each detection station; the attitude adjustment module is used to adjust the shooting angle of the vision module by moving the vertical axis, pitch axis and rotation axis; the vision module is used to capture images of the machine parts and has focusing and multi-angle lighting functions. The overall control system is connected to the posture adjustment module and the vision module in the vision inspection device, respectively, and is used to control the posture adjustment module and the vision module. The overall control system displays images of the captured body parts and performs functions such as manual selection, position correction, contour recognition, contour comparison and detection database recording. The overall control system has a calibration mode or a detection mode. In calibration mode, it works with a vision inspection device to acquire the detection template for each detection station for a machine body component with correctly installed corner pieces, and forms a detection template database by combining the layout of all detection stations. In detection mode, it works with a vision inspection device to match the images of the machine body component to be inspected taken at each detection station with the detection templates at the corresponding detection stations to obtain the corner piece detection results for each detection station.

10. The method according to claim 9, characterized in that, The overall control system has the functions of calling the detection template database, controlling the posture adjustment module to adjust the posture, controlling the vision module to capture images and complete the image-based error and omission identification and analysis, and storing and analyzing the detection results.

Images