A workpiece blank edge precision detection system and method based on machine vision

Abstract

This application discloses a machine vision-based precision inspection system and method for workpiece blank edges. The system includes a transport unit; a rotation unit; a detection area on the rotating disk; a CCD unit; a feeding unit; a feeding mechanism electrically connected to the CCD unit; and a laser engraving and marking unit. After the workpiece is moved to the transport mechanism via a vibrating disk and outlined by the laser engraving and marking unit, a guiding mechanism pushes the workpiece from the rotating disk to the detection area. After being inspected by a first detection mechanism and a second detection mechanism, the workpiece is moved to the corresponding storage box by the feeding mechanism. This application also discloses a machine vision-based precision inspection method for workpiece blank edges, including a machine vision-based precision inspection system for workpiece blank edges, which integrates automatic feeding, transfer, multi-directional visual inspection, and automatic sorting functions, realizing full-process automation and improving inspection efficiency and accuracy.

Description

Technical Field

[0001] This application relates to the fields of industrial machine vision and photoelectric inspection technology, and in particular to a precision inspection system and method for workpiece blank edges based on machine vision. Background Technology

[0002] In the manufacturing process of precision metal workpieces, such as connector terminals, precision stampings, and machined parts, the integrity of the workpiece edges (i.e., blank edges) is one of the key indicators for measuring their quality. Blank edge defects, such as burrs, notches, collapsed edges, and rolled edges, not only affect the appearance and dimensional accuracy of the workpiece but may also lead to failures in subsequent assembly, electrical performance, or mechanical properties. Therefore, inspecting the workpiece blank edges is an essential step in ensuring product quality.

[0003] Currently, to overcome the shortcomings of manual inspection, automated inspection equipment using CCD cameras has emerged on the market, thereby testing or analyzing the material's quality based on the workpiece's physical properties. However, this automated inspection equipment still has many limitations when applied to the detection of defects on workpiece blank edges: Limited inspection scope: Most equipment only photographs the top surface of the workpiece from a fixed angle, making it difficult to effectively capture defects on the side edges of the workpiece. To achieve multi-side inspection, multiple camera stations or complex flipping mechanisms are required, resulting in bulky equipment structures, high costs, and difficulty in guaranteeing the positioning accuracy between stations, affecting the consistency of inspection.

[0004] Low level of automation integration: The feeding, inspection, and sorting processes are often relatively independent and lack smooth automation connections. For example, feeding still requires manual placement, and sorting of qualified and unqualified products requires manual intervention or is completed by simple mechanical pushers, making it impossible to achieve automatic online inspection and classification. Summary of the Invention

[0005] To address the shortcomings of limited detection dimensions and low automation integration, this application provides a machine vision-based precision inspection system for workpiece blank edges.

[0006] The workpiece blank edge precision inspection system based on machine vision provided in this application adopts the following technical solution: A machine vision-based precision inspection system for workpiece blank edges, including A transport unit, comprising a vibratory feeder and a transmission mechanism in communication with the vibratory feeder; The rotating unit includes a machine base, a rotating disk mounted on the machine base, and a material guiding mechanism; the rotating disk is provided with a detection area. The CCD unit includes a first detection mechanism for detecting the side surface of the workpiece and a second detection mechanism for detecting the top surface of the workpiece. The feeding unit includes a storage box and a feeding mechanism corresponding to the storage box; the feeding mechanism is electrically connected to the CCD unit. A laser engraving marking unit is located on the transmission path of the transmission mechanism and is used to outline the edge contour of the workpiece that passes through it. After the workpiece is moved to the transmission mechanism by the vibrating plate and outlined by the laser engraving marking unit, the guiding mechanism pushes the workpiece on the rotating plate to the detection area. After being detected by the first detection mechanism and the second detection mechanism, the workpiece is moved to the corresponding storage box by the unloading mechanism.

[0007] By adopting the above technical solution, automatic feeding, transfer, multi-directional visual inspection and automatic sorting functions are integrated, realizing the full-process automation of workpieces from feeding to classification output, reducing manual intervention, and actively processing high-contrast permanent marking lines on the edge of the workpiece before inspection, providing stable and clear contour features for subsequent image recognition, significantly reducing the difficulty of edge recognition caused by the workpiece's original color, reflection or uneven lighting, and improving the accuracy of measurement.

[0008] Preferably, the material guiding mechanism includes a material guiding motor and a material guiding turntable disposed on the material guiding motor. After the workpiece falls into the predetermined area, it moves with the turntable toward the material guiding turntable and is pushed by the material guiding turntable to the detection area located on the edge of the turntable.

[0009] By adopting the above technical solution, the rotation of the rotating disk and the active pushing of the guide turntable can be combined to smoothly push the workpiece at any angle into the detection area. After entering the detection area, the workpiece will continue to move for the CCD unit to acquire images.

[0010] Preferably, the CCD unit further includes a first supplementary lighting mechanism corresponding to the first detection mechanism and a second supplementary lighting mechanism corresponding to the second detection mechanism. The first detection mechanism and the first supplementary lighting mechanism are located on the inner and outer sides of the detection area, and the second detection mechanism and the second supplementary lighting mechanism are located on the upper and lower sides of the detection area.

[0011] By adopting the above technical solution, dedicated supplementary lighting sources are configured for the side and top surface inspection of the workpiece, which can optimize the lighting conditions for different shooting directions, effectively suppress interference such as shadows and reflections, and ensure that the acquired workpiece image features are clear and have high contrast, thereby improving the accuracy of defect identification.

[0012] Preferably, the machine base is further provided with a position adjustment mechanism, which is located between the material guiding mechanism, the first detection mechanism, the second detection mechanism, the first supplementary lighting mechanism and the second supplementary lighting mechanism and the machine base.

[0013] By adopting the above technical solutions, the positions of key functional components can be flexibly adjusted according to the size, shape, or testing requirements of different workpieces, thereby enhancing the equipment's adaptability to different types of workpieces and improving its versatility and utilization.

[0014] Preferably, the position adjustment mechanism includes a transverse track on the machine base, a transverse slide rail on the transverse track, a transverse moving platform on the transverse slide rail, a lifting slide rail on the transverse moving platform, a lifting platform on the lifting slide rail, a transverse locking member for restricting the movement of the transverse moving platform, and a lifting locking member for restricting the movement of the lifting platform. Any one of the material guiding mechanism, the first detection mechanism, the second detection mechanism, the first supplementary lighting mechanism, and the second supplementary lighting mechanism is installed on the lifting platform.

[0015] By adopting the above technical solution, the ability to make independent and precise position adjustments in both horizontal and vertical dimensions is provided, and the locking mechanism ensures that the adjusted position is stable, enabling more precise alignment and focusing, and meeting the requirements of high-precision detection.

[0016] Preferably, the storage box includes a good product box, a defective product box, and a box to be inspected. The unloading mechanism includes a first air valve, a second air valve, and a third air valve arranged toward the inspection area. The first air valve, the second air valve, and the third air valve are used to blow the workpieces into the corresponding good product box, the defective product box, and the box to be inspected.

[0017] By adopting the above technical solution, the unloading mechanism uses compressed air as power to achieve non-contact rapid sorting based on the test results, accurately blowing the workpieces into the corresponding collection containers. The action is rapid and there is no risk of mechanical contact damage. At the same time, the workpieces to be inspected are independently stored, which is convenient for subsequent review and testing.

[0018] Preferably, it also includes a positioning mechanism, which is used to limit the workpiece on the transmission mechanism and release the workpiece after the outline processing, so that the workpiece falls into the rotating disk.

[0019] By adopting the above technical solution, the workpiece is kept stationary during the laser engraving process, thereby obtaining a mark with accurate position and uniform lines; after the engraving is completed, it is automatically released, allowing the workpiece to smoothly enter the next process, realizing a reliable connection between laser engraving and the inspection station and process automation.

[0020] Preferably, the laser engraving marking unit includes a horizontal movement mechanism, a vertical movement mechanism disposed on the horizontal movement mechanism, a lifting mechanism disposed on the vertical movement mechanism, a steering mechanism disposed on the lifting mechanism, a tilting mechanism disposed on the steering mechanism, and a laser engraving machine disposed on the tilting mechanism, wherein the tilting mechanism is used to adjust the angle between the laser engraving machine and the horizontal plane.

[0021] The laser marking unit has strong adaptability and can quickly adjust to the optimal processing posture for workpieces of different heights, positions and with complex contours, ensuring the consistency and high contrast of the marking lines.

[0022] This invention also provides a machine vision-based method for precision inspection of workpiece blank edges, employing the machine vision-based precision inspection system for workpiece blank edges as described above, including the following steps: S1: Feeding and conveying, the metal workpieces are sorted one by one by the vibratory feeder and output to the conveying mechanism; S2: Outlining process: The workpiece passes through the laser engraving station on the transmission mechanism, and the laser engraving marking unit outlines the edge contour of the workpiece to form a marking line. S3: Positioning and loading. The conveying mechanism transports the workpiece to the predetermined area of ​​the rotating disk, and the guiding mechanism pushes the workpiece to the detection area on the rotating disk. S4: Multi-directional image acquisition. The CCD unit is activated to capture images of the workpiece that has undergone outline processing. The first detection mechanism is used to acquire side images of the workpiece, and the second detection mechanism is used to acquire top images of the workpiece. The top or side images contain the marking lines. S5: Defect judgment and classification. Based on the acquired side and top images, analyze whether there are blank edge defects in the workpiece; generate corresponding classification signals based on the analysis results. S6: Automatic sorting and unloading: Based on the classification signal, control the unloading mechanism to sort the workpieces and transfer them to the corresponding storage boxes.

[0023] By adopting the above technical solutions, the automated equipment is combined with the systematic process flow, realizing online, continuous, multi-angle automatic detection and real-time sorting of defects on the edge of metal workpiece blanks, forming a complete closed-loop quality control system, which greatly improves the overall efficiency and reliability of the inspection operation.

[0024] Preferably, in step S5, the detected image is processed into a grid, divided into multiple pixel regions, and the contour regions formed by the marker lines in the image are identified; the actual pixel width or area of ​​the contour regions is calculated and compared with a preset standard pixel range; If the actual pixel value does not fall within the standard pixel range, but the deviation from the standard pixel range is within the preset allowable tolerance range, then the workpiece is determined to have a blank edge defect and is a defective part. If the actual pixel value is within the standard pixel range, the workpiece is determined to be free of blank edge defects and is a good product. If the actual pixel value does not fall within the standard pixel range and the deviation from the standard pixel range exceeds the allowable error range, the detection result is deemed invalid and the workpiece is to be inspected.

[0025] By adopting the above technical solution, edge defects that are difficult to quantify manually are transformed into quantitative calculations and comparisons of image pixels. By setting allowable tolerance ranges, the degree of defects can be finely graded and judged, as well as abnormal situations can be detected, thereby improving the accuracy and scientific nature of quality control.

[0026] In summary, this application includes at least one of the following beneficial technical effects: 1. Fully automated, multi-dimensional, high-precision online inspection has been achieved. Through the coordination of the vibrating plate, transmission mechanism, rotating plate, and guiding mechanism, automatic feeding, precise positioning, and transfer of workpieces are realized. By using the first and second inspection mechanisms to acquire images from the side and top surfaces respectively, combined with special supplementary lighting, multi-angle, high-definition inspection of defects on the workpiece blank edge is achieved, overcoming the limitations of traditional single-angle inspection and significantly improving the defect detection rate and inspection efficiency.

[0027] 2. Improved equipment adaptability and enhanced objectivity and accuracy of inspection. The position adjustment mechanism allows for adjustable positions of key components, facilitating rapid adaptation to workpieces of different specifications. The laser marking unit pre-creates high-contrast marking lines on the workpiece edge, and then, through image gridding processing and pixel quantitative calculation, transforms traditional experience-based defect judgment into objective, quantitative data comparison, eliminating subjective errors from manual visual inspection and ensuring consistency of inspection standards and accuracy of results. 3. A complete intelligent quality sorting closed loop has been formed. The detection system and the unloading mechanism are directly linked, and the corresponding air valves can be controlled in real time based on image analysis results to automatically sort the workpieces into good, defective, and inspection boxes. This achieves unmanned operation of detection and sorting, improving production efficiency. By sorting the parts to be inspected, the isolation and management of measurement anomalies are achieved, providing support for production quality traceability and process improvement. Attached Figure Description

[0028] Figure 1 This is a three-dimensional structural diagram of an embodiment of this application.

[0029] Figure 2 yes Figure 1 Internal structure diagram Figure 1.

[0030] Figure 3 yes Figure 1 Internal structure diagram Figure 2 .

[0031] Figure 4 This is a schematic diagram of the positioning mechanism.

[0032] Figure 5 This is a schematic diagram of the structure of the laser marking unit.

[0033] Figure 6 This is a flowchart of a machine vision-based precision inspection method for workpiece blank edges according to this application.

[0034] Explanation of reference numerals in the attached figures: 1. Transport unit; 11. Vibratory feeder; 12. Transmission mechanism; 121. Base; 122. Transmission track; 2. Rotation unit; 21. Machine platform; 22. Rotary disk; 23. Material guiding mechanism; 231. Material guiding motor; 232. Material guiding turntable; 24. Detection area; 25. Predetermined area; 3. CCD unit; 31. First detection mechanism; 32. Second detection mechanism; 33. First supplementary lighting mechanism; 34. Second supplementary lighting mechanism; 4. Unloading unit; 41. Storage box; 411. Good product box; 412. Defective product box; 413. Box to be inspected; 42. Unloading mechanism; 421. First air blowing valve; 42 2. Second air valve; 423. Third air valve; 424. Fourth air valve; 43. Main frame; 44. Sub-frame; 5. Position adjustment mechanism; 51. Transverse track; 52. Transverse slide rail; 53. Transverse platform; 54. Lifting slide rail; 55. Lifting platform; 56. Transverse locking component; 57. Lifting locking component; 6. Laser engraving marking unit; 61. Transverse mechanism; 62. Longitudinal mechanism; 63. Lifting mechanism; 64. Steering mechanism; 65. Tilt mechanism; 66. Laser engraving machine; 7. Positioning mechanism; 71. Telescopic cylinder; 72. Positioning plate; 721. Interception part; 722. Positioning part; 8. Workpiece. Detailed Implementation

[0035] The following is in conjunction with the appendix Figures 1 to 6 This application will be described in further detail.

[0036] This application discloses a machine vision-based precision inspection system for workpiece blank edges, referring to... Figure 1 ,include The transport unit 1 includes a vibratory feeder 11 and a transmission mechanism 12 connected to the vibratory feeder 11; The rotating unit 2 includes a machine base 21, a rotating disk 22 disposed on the machine base 21, and a material guiding mechanism 23; the rotating disk 22 is provided with a detection area 24; wherein the rotating disk is made of transparent material.

[0037] CCD unit 3 includes a first detection mechanism 31 for detecting the side surface of the workpiece and a second detection mechanism 32 for detecting the top surface of the workpiece. The feeding unit 4 includes a storage box 41 and a feeding mechanism 42 corresponding to the storage box 41; the feeding mechanism 42 is electrically connected to the CCD unit 3. The laser engraving marking unit 6 is located on the transmission path of the transmission mechanism 12 and is used to outline the edge contour of the workpiece that passes through it. After the workpiece 8 is moved to the transmission mechanism 12 by the vibrating plate 11 and is outlined by the laser marking unit, the guiding mechanism 23 pushes the workpiece on the rotating plate 22 to the detection area 24. After the workpiece is detected by the first detection mechanism 31 and the second detection mechanism 32, it is moved to the corresponding storage box 41 by the unloading mechanism 42.

[0038] The vibratory feeder 11 acts as an automatic feeding source, orienting and sorting the scattered workpieces 8 and outputting them through its outlet. The conveying mechanism 12 receives the workpieces from the vibratory feeder and smoothly transports them to a predetermined area. During the rotation of the rotating disc, the guiding mechanism 23 pushes the workpieces from the initial predetermined area 25 to a fixed detection area 24. As the rotating disc rotates, the workpieces move to the position corresponding to the CCD unit 3, which acquires images from the side and top. The image processing system analyzes the acquired images and makes a quality judgment. This judgment signal controls the corresponding air valve of the unloading unit 4 to separate and blow the workpieces into different storage boxes 41. This solution integrates automatic feeding, precise positioning, multi-angle imaging, and intelligent sorting functions, realizing a fully automated closed-loop process from disordered workpieces to classified output, significantly improving detection efficiency and objectivity, and reducing manual intervention and its associated errors.

[0039] Furthermore, referring to Figure 2 The material guiding mechanism 23 includes a material guiding motor 231 and a material guiding turntable 232 disposed on the material guiding motor 231. After the workpiece falls into the predetermined area 25, it moves with the rotating disk 22 toward the material guiding turntable 232 and is pushed by the material guiding turntable 232 to the detection area 24 located on the edge of the rotating disk 22.

[0040] The guide motor 231 drives the guide turntable 232 to rotate continuously or as needed. When the turntable 22, carrying the workpiece 8 that has fallen into the predetermined area 25, rotates to a position close to the guide turntable 232, the rotating guide turntable 232 generates a continuous tangential thrust through contact with the side of the workpiece, smoothly moving and guiding the workpiece gradually from the predetermined area 25 inside the turntable until it is accurately positioned and limited to the fixed detection area 24 on the circumferential edge of the turntable 22. This achieves dynamic and flexible mechanical positioning, avoids rigid impact, and ensures that the workpiece enters the detection area with a consistent and stable posture, providing a reliable positional reference for subsequent high-precision and high-repeatability visual inspection. In this embodiment, the detection area is the edge position of the turntable.

[0041] Furthermore, referring to Figure 2 The CCD unit 3 further includes a first supplementary light mechanism 33 corresponding to the first detection mechanism 31 and a second supplementary light mechanism 34 corresponding to the second detection mechanism 32. The first detection mechanism 31 and the first supplementary light mechanism 33 are located on the inner and outer sides of the detection area 24, and the second detection mechanism 32 and the second supplementary light mechanism 34 are located on the upper and lower sides of the detection area 24.

[0042] The first detection mechanism 31 and the first supplementary lighting mechanism 33 are respectively arranged on the inner and outer sides of the detection area 24, forming a side imaging light path for capturing the sidewall and edge contours of the workpiece. The second detection mechanism 32 and the second supplementary lighting mechanism 34 are located above and below the detection area 24, forming a vertical imaging light path for acquiring the top surface features of the workpiece. The two systems can be triggered sequentially to acquire images under optimal lighting conditions. By independently configuring dedicated light sources for cameras with different viewing angles, the lighting conditions of the side and top surfaces are specifically optimized, effectively overcoming interference from shadows and reflections, thereby significantly improving the overall image quality, contrast, and feature clarity, and enhancing the accuracy of defect identification.

[0043] Furthermore, referring to Figure 2 The machine base 21 is also equipped with a position adjustment mechanism 5, which is located between the machine base 21 and any one of the following: the material guiding mechanism 23, the first detection mechanism 31, the second detection mechanism 32, the first supplementary lighting mechanism 33, and the second supplementary lighting mechanism 34. The position adjustment mechanism 5 acts as an adjustable installation intermediary, positioned between the machine base 21 and the component requiring adjustment, such as the first detection mechanism 31. When the position of this component needs to be adjusted, the corresponding adjustment and locking components of the position adjustment mechanism 5 can be operated to cause the component to move the required distance relative to the fixed detection area 24, and then locked after adjustment. This provides position adjustment capability, enabling the equipment to quickly adapt to the detection needs of workpieces of different sizes and shapes, significantly enhancing the versatility and production flexibility of the equipment, and shortening the adjustment time during product changeovers.

[0044] Furthermore, referring to Figure 3 The position adjustment mechanism 5 includes a transverse rail 51 on the machine base 21, a transverse slide rail 52 on the transverse rail 51, a transverse moving platform 53 on the transverse slide rail 52, a lifting slide rail 54 on the transverse moving platform 53, a lifting platform 55 on the lifting slide rail 54, a transverse locking member 56 for restricting the movement of the transverse moving platform 53, and a lifting locking member 57 for restricting the movement of the lifting platform 55. Any one of the material guiding mechanism 23, the first detection mechanism 31, the second detection mechanism 32, the first supplementary lighting mechanism 33, and the second supplementary lighting mechanism 34 is mounted on the lifting platform 55. The transverse moving platform 53 can slide along the transverse slide rail 52 on the transverse rail 51 to achieve coarse or fine adjustment of the horizontal position, and is locked with the transverse locking member 56 after adjustment. Secondly, the lifting platform 55 can move up and down along the lifting slide rail 54 mounted on the transverse moving platform 53 to achieve vertical position adjustment, and is locked with the lifting locking member 57 after adjustment. The components to be adjusted are mounted on the lifting platform 55. Through precise mechanical guide rails and locking structures, independent, stable, and accurate two-dimensional position adjustment and fixation of components such as cameras and light sources in the horizontal and vertical directions are achieved, meeting the requirements of high-precision vision systems for installation position and angle.

[0045] Furthermore, referring to Figure 3 The storage box 41 includes a good product box 411, a defective product box 412, and a box to be inspected 413. The unloading mechanism 42 includes a first air valve 421, a second air valve 422, and a third air valve 423 arranged towards the detection area 24. The first air valve 421, the second air valve 422, and the third air valve 423 are used to blow the workpieces into the corresponding good product box 411, the defective product box 412, and the box to be inspected 413. Based on the image analysis results of the CCD unit 3, the workpiece is determined to be "good," "defective," or "to be inspected." When the rotating disk 22 transfers the inspected workpiece to the unloading station, the control system triggers the corresponding first air valve 421, second air valve 422, or third air valve 423 to open instantaneously according to the determined category. Compressed air forms a directional airflow, blowing the workpieces off the rotating disk 22 and causing them to fall along a preset trajectory into the good product box 411, defective product box 412, or inspection box 413 below. This achieves instantaneous, non-contact automatic sorting based on the inspection results. The operation is fast and without mechanical wear, avoiding surface damage to the workpieces that may be caused by traditional robotic arm sorting. The "inspection box" isolates suspicious or failed-inspection workpieces for subsequent verification.

[0046] Furthermore, referring to Figure 2The feeding mechanism 42 also includes a fourth air-blowing valve 424, which is used to blow workpieces on the rotating disk 22 that are outside the detection area 24 or the predetermined area 25 off the rotating disk 22. When the system detects or anticipates that a workpiece 8 may accidentally remain in the non-working area of ​​the rotating disk 22, or during periodic cleaning, the fourth air-blowing valve 424 is activated, spraying air to blow the stranded workpiece off the rotating disk, allowing it to enter the waste trough or recycling area. This provides an active cleaning function, effectively removing abnormal residual workpieces from the rotating disk, preventing equipment malfunctions, false detections, or damage caused by jamming or stacking, and greatly improving the long-term stability and reliability of the equipment.

[0047] Furthermore, referring to Figure 2 A main frame 43 for adjusting the position of the fourth air valve 424 is provided between the fourth air valve 424 and the lifting platform 55. The fourth air valve 424 is mounted on the adjustable lifting platform 55 or other fixed structure via the main frame 43. The main frame 43 itself can be designed to have angle or position fine adjustment functions to ensure that the airflow of the fourth air valve 424 can be accurately aimed at the area to be cleaned.

[0048] Furthermore, referring to Figure 2 The fourth air-blowing valve 424 is located at the center of the rotating disk 22. It ensures that the airflow direction is towards the outside of the rotating disk, so that the airflow blows the stuck workpiece off the outside of the rotating disk.

[0049] Furthermore, referring to Figure 3 The main frame 43 is equipped with a sub-frame 44 for adjusting the first supplementary lighting mechanism 33. The sub-frame 44, as an additional mounting structure on the main frame 43, is used to fix and adjust the position and angle of the first supplementary lighting mechanism 33. The addition of the main frame 43 and the sub-frame 44 allows for independent and flexible adjustment of the positions of auxiliary components such as the fourth air valve 424 and the first supplementary lighting mechanism 33, further optimizing the cleaning and lighting effects and enhancing the overall system's ease of commissioning and adaptability.

[0050] Furthermore, referring to Figure 5The laser engraving marking unit 6 includes a horizontal movement mechanism 61, a vertical movement mechanism 62 mounted on the horizontal movement mechanism, a lifting mechanism 63 mounted on the vertical movement mechanism 62, a steering mechanism 64 mounted on the lifting mechanism, a tilting mechanism 65 mounted on the steering mechanism, and a laser engraving machine 66 mounted on the tilting mechanism 65. The tilting mechanism 65 is used to adjust the angle between the laser engraving machine 66 and the horizontal plane. The horizontal movement mechanism 61, the vertical movement mechanism 62, and the lifting mechanism 63 are responsible for adjusting the approximate position of the laser engraving machine 66 in three-dimensional space. The steering mechanism 64 and the tilting mechanism 65 are used to fine-tune the horizontal azimuth angle and vertical tilt angle of the laser emitter head, ensuring that the laser beam is always perpendicular to the cut surface of the workpiece 8 to be engraved, adapting to workpieces of different shapes, and perpendicular to any surface on the workpiece that needs precise detection. Thus, high-quality marking lines of uniform depth and width can be etched under any workpiece posture.

[0051] Furthermore, referring to Figure 4 The transmission mechanism 12 includes a base 121 and a transmission track 122 disposed on the base 121 and connected to the discharge port of the vibrating plate.

[0052] Furthermore, referring to Figure 4 It also includes a positioning mechanism 7, which limits the workpiece on the transmission mechanism 12 and releases the limit after outlining, allowing the workpiece to fall into the rotating disk 22. The transmission mechanism 12 is stably supported by a base 121. The transmission track 122 is fixed to the base 121, with one end precisely aligned with the discharge port of the vibratory disk 11 to receive the sorted workpieces 8. The structure is simple and reliable, achieving a smooth and orderly transition from the vibratory disk to the processing / inspection station, ensuring that the posture of the workpiece remains basically unchanged during the conveying process, and improving the accuracy of the transmission.

[0053] Furthermore, referring to Figure 4 The positioning mechanism 7 includes a telescopic cylinder 71 corresponding to the transmission track 122, and a positioning plate 72 disposed on the output end of the telescopic cylinder 71. The positioning plate includes an intercepting part 721, on which a first magnet is disposed. When the output end of the telescopic cylinder extends, it drives the positioning plate to the positioning position, and the intercepting part abuts against the outlet of the transmission track to limit the workpiece from falling. The first magnet attracts the workpiece closest to the outlet, i.e., the first workpiece, to limit the movement of the workpiece.

[0054] Furthermore, referring to Figure 4The positioning plate also includes a positioning part 722 connected to the intercepting part in an L-shape. The positioning part is equipped with a second magnet, which is used to attract the second workpiece, maintaining a distance between the second and first workpieces to prevent subsequent workpieces from affecting the positioning of the first workpiece. After the first workpiece undergoes outlining, when the output end of the telescopic cylinder retracts, it moves the positioning plate to the release position, the intercepting part leaves the transmission track outlet, and the first workpiece falls into the predetermined area. At this time, the second workpiece also comes into contact with the second magnet and continues to move. Because there is a distance between the first and second workpieces, the telescopic cylinder has sufficient time to respond and extend to intercept the second workpiece after the first workpiece falls into the predetermined area. This process is repeated. This achieves precise positioning and fixation of a single workpiece during laser engraving. Furthermore, through magnetic adsorption spacing control, it solves the problem of workpiece queuing and spacing in continuous feeding, ensuring seamless and efficient connection between single-piece processing and continuous operation, effectively preventing workpiece accumulation or collision, and improving the overall production cycle time.

[0055] This embodiment also provides a machine vision-based method for precision inspection of workpiece blank edges, referencing... Figure 6 The above-described machine vision-based precision inspection system for workpiece blank edges includes the following steps: S1: Feeding and conveying, the metal workpieces are sorted one by one by the vibratory feeder and output to the conveying mechanism; S2: Outlining process: The workpiece passes through the laser engraving station on the transmission mechanism, and the laser engraving marking unit outlines the edge contour of the workpiece to form a marking line. S3: Positioning and loading. The conveying mechanism transports the workpiece to the predetermined area of ​​the rotating disk, and the guiding mechanism pushes the workpiece to the detection area on the rotating disk. S4: Multi-directional image acquisition. The CCD unit is activated to capture images of the workpiece that has undergone outline processing. The first detection mechanism is used to acquire side images of the workpiece, and the second detection mechanism is used to acquire top images of the workpiece. The top or side images contain the marking lines. S5: Defect judgment and classification. Based on the acquired side and top images, analyze whether there are blank edge defects in the workpiece; generate corresponding classification signals based on the analysis results. S6: Automatic sorting and unloading: Based on the classification signal, control the unloading mechanism to sort the workpieces and transfer them to the corresponding storage boxes.

[0056] By combining automated equipment with a systematic process, online, continuous, multi-angle automatic detection and real-time sorting of defects on the edges of metal workpieces are achieved, forming a complete closed-loop quality control system that significantly improves the overall efficiency and reliability of the inspection operation.

[0057] Further, in step S5, the detected image is processed into a grid, divided into multiple pixel regions, and the contour regions formed by the marker lines in the image are identified; the actual pixel width or area of ​​the contour regions is calculated and compared with a preset standard pixel range; If the actual pixel value does not fall within the standard pixel range, but the deviation from the standard pixel range is within the preset allowable tolerance range, then the workpiece is determined to have a blank edge defect and is a defective part. If the actual pixel value is within the standard pixel range, the workpiece is determined to be free of blank edge defects and is a good product. If the actual pixel value does not fall within the standard pixel range and the deviation from the standard pixel range exceeds the allowable error range, the detection result is deemed invalid and the workpiece is to be inspected.

[0058] By adopting the above technical solution, edge defects that are difficult to quantify manually are transformed into quantitative calculations and comparisons of image pixels. By setting allowable tolerance ranges, the degree of defects can be finely graded and judged, as well as abnormal situations can be detected, thereby improving the accuracy and scientific nature of quality control.

[0059] The implementation principle of a machine vision-based precision inspection system for workpiece blank edges in this application is as follows: First, the vibratory feeder 11 orients and sorts the disordered workpieces 8 through vibration, and outputs them one by one to the transmission track 122 of the transmission mechanism 12. The workpieces are conveyed along the track. When they pass the laser engraving station, the telescopic cylinder 71 of the positioning mechanism 7 drives the positioning plate 72 to extend. Its intercepting part 721 precisely intercepts and fixes the first workpiece to be processed, while the positioning part 722 attracts the second workpiece to maintain the spacing. The laser engraving marking unit 6 then starts, ensuring that the laser head quickly etches a continuous, high-contrast marking line on the edge contour of the first workpiece. After completion, the positioning mechanism 7 resets, the first workpiece falls into the predetermined area 25 of the rotating disk 22, the second workpiece moves forward, and the cycle repeats. The rotating disk 22 rotates, and the guiding mechanism 23 on it, through the rotation of the guiding turntable 232, smoothly and accurately pushes the workpiece from the predetermined area 25 to the fixed detection area 24.

[0060] When the workpiece is stably positioned in the detection area 24, the CCD unit 3 is activated. Under the side illumination of the first supplementary lighting mechanism 33, the first detection mechanism 31 acquires a side image of the workpiece; under the bottom illumination of the second supplementary lighting mechanism 34, the second detection mechanism 32 acquires a top image of the workpiece. The core feature of the top image is the high-contrast marking lines pre-made by laser engraving, which clearly outline the actual edge contour of the workpiece, providing stable and reliable visual features for subsequent analysis. Secondly, during the laser engraving process, impurities such as burrs and debris on the workpiece surface can be removed, especially burrs that exceed the acceptable dimensions of the workpiece.

[0061] After preprocessing the acquired inspection image, the image processing system performs gridding, dividing it into a regular pixel grid to facilitate digital analysis by the computer. Subsequently, edge detection and contour extraction algorithms accurately identify the contour regions formed by laser-engraved marking lines in the image. The system calculates the actual pixel width or pixel area of ​​this contour region and quantifies it into a specific value. This value is compared with a pre-calibrated standard pixel range using a defect-free standard part. The system introduces an allowable tolerance range as a buffer criterion: if the measured value is within the standard pixel range, it is judged as a good product; if it exceeds the standard range but is within the allowable tolerance, it is judged as a defective product with acceptable defects; if it exceeds the allowable tolerance, the inspection result is judged as abnormal, such as a serious defect or image malfunction, and it is classified as a product to be inspected. This process transforms subjective visual judgment into objective, quantifiable data-driven decisions.

[0062] Based on the classification signals "good product," "defective product," and "to be inspected" generated in step S5, when the workpiece rotates with the rotating disk 22 to the unloading station, the corresponding execution unit of the unloading mechanism 42 is triggered: the first air blowing valve 421, the second air blowing valve 422, or the third air blowing valve 423. Compressed air generates a directional pulsed airflow, which non-contactly blows the workpiece away from the rotating disk and makes it fall accurately into the corresponding good product box 411, defective product box 412, or to be inspected box 413, thereby completing the physical closed loop from "inspection" to "sorting."

[0063] The above are all preferred embodiments of this application, and are not intended to limit the scope of protection of this application. Therefore, all equivalent changes made in accordance with the structure, shape and principle of this application should be covered within the scope of protection of this application.

Claims

1. A machine vision-based precision detection system for workpiece blank edge, characterized in that: include The transport unit (1) includes a vibrating plate (11) and a transmission mechanism (12) connected to the vibrating plate (11); The rotating unit (2) includes a machine base (21), a rotating disk (22) disposed on the machine base (21), and a material guiding mechanism (23); the rotating disk (22) is provided with a detection area (24); CCD unit (3) includes a first detection mechanism (31) for detecting the side surface of the workpiece and a second detection mechanism (32) for detecting the top surface of the workpiece. The unloading unit (4) includes a storage box (41) and an unloading mechanism (42) corresponding to the storage box (41); the unloading mechanism (42) is electrically connected to the CCD unit (3); A laser engraving marking unit (6) is provided on the transmission path of the transmission mechanism (12) and is used to outline the edge contour of the workpiece that passes through. The workpiece is moved to the transmission mechanism (12) by the vibrating plate (11) and after being outlined by the laser marking unit, it is moved one by one to the predetermined area (25) of the rotating plate (22). The guiding mechanism (23) is used to push the workpiece on the rotating plate (22) to the detection area (24). After the workpiece is detected by the first detection mechanism (31) and the second detection mechanism (32), it is moved to the corresponding storage box (41) by the unloading mechanism (42).

2. The machine vision-based precision inspection system for workpiece blank edges according to claim 1, characterized in that: The material guiding mechanism (23) includes a material guiding motor (231) and a material guiding turntable (232) disposed on the material guiding motor (231). After the workpiece falls into the predetermined area (25), it moves with the rotating disk (22) toward the material guiding turntable (232) and is pushed by the material guiding turntable (232) to the detection area (24) located on the edge of the rotating disk (22).

3. The machine vision-based precision inspection system for workpiece blank edges according to claim 1, characterized in that: The CCD unit (3) further includes a first supplementary light mechanism (33) corresponding to the first detection mechanism (31) and a second supplementary light mechanism (34) corresponding to the second detection mechanism (32). The first detection mechanism (31) and the first supplementary light mechanism (33) are located on the inner and outer sides of the detection area (24), and the second detection mechanism (32) and the second supplementary light mechanism (34) are located on the upper and lower sides of the detection area (24).

4. The machine vision-based precision inspection system for workpiece blank edges according to claim 3, characterized in that: The machine base (21) is also provided with a position adjustment mechanism (5), which is located between the machine base (21) and any one of the material guiding mechanism (23), the first detection mechanism (31), the second detection mechanism (32), the first supplementary lighting mechanism (33) and the second supplementary lighting mechanism (34).

5. The machine vision-based precision inspection system for workpiece blank edges according to claim 4, characterized in that: The position adjustment mechanism (5) includes a transverse track (51) on the machine base (21), a transverse slide rail (52) on the transverse track (51), a transverse moving platform (53) on the transverse slide rail (52), a lifting slide rail (54) on the transverse moving platform (53), a lifting platform (55) on the lifting slide rail (54), a transverse moving locking member (56) for restricting the movement of the transverse moving platform (53), and a lifting locking member (57) for restricting the movement of the lifting platform (55). Any one of the material guiding mechanism (23), the first detection mechanism (31), the second detection mechanism (32), the first supplementary lighting mechanism (33), and the second supplementary lighting mechanism (34) is installed on the lifting platform (55).

6. The machine vision-based precision inspection system for workpiece blank edges according to claim 1, characterized in that: The storage box (41) includes a good product box (411), a defective product box (412), and a box to be inspected (413). The unloading mechanism (42) includes a first air valve (421), a second air valve (422), and a third air valve (423) arranged toward the detection area (24). The first air valve (421), the second air valve (422), and the third air valve (423) are used to blow the workpieces into the corresponding good product box (411), the defective product box (412), and the box to be inspected (413).

7. The machine vision-based precision inspection system for workpiece blank edges according to claim 1, characterized in that: It also includes a positioning mechanism (7), which is used to limit the workpiece on the transmission mechanism (12) and release the workpiece after the outline processing so that the workpiece falls into the rotating disk (22).

8. The machine vision-based precision inspection system for workpiece blank edges according to claim 1, characterized in that: The laser marking unit (6) includes a horizontal movement mechanism (61), a vertical movement mechanism (62) provided on the horizontal movement mechanism (61), a lifting mechanism (63) provided on the vertical movement mechanism (62), a steering mechanism (64) provided on the lifting mechanism (63), a tilting mechanism (65) provided on the steering mechanism (64), and a laser engraving machine (66) provided on the tilting mechanism (65). The tilting mechanism (65) is used to adjust the angle between the laser engraving machine (66) and the horizontal plane.

9. A method for precision inspection of workpiece blank edges based on machine vision, characterized in that: The machine vision-based precision inspection system for workpiece blank edges, as described in any one of claims 1 to 8, includes the following steps: S1: Feeding and conveying, the metal workpieces are sorted one by one by the vibratory feeder and output to the conveying mechanism; S2: Outlining process: The workpiece passes through the laser engraving station on the transmission mechanism, and the laser engraving marking unit outlines the edge contour of the workpiece to form a marking line. S3: Positioning and loading. The conveying mechanism transports the workpiece to the predetermined area of ​​the rotating disk, and the guiding mechanism pushes the workpiece to the detection area on the rotating disk. S4: Multi-directional image acquisition. The CCD unit is activated to capture images of the workpiece that has undergone outline processing. The first detection mechanism is used to acquire side images of the workpiece, and the second detection mechanism is used to acquire top images of the workpiece. The top or side images contain the marking lines. S5: Defect judgment and classification. Based on the acquired side and top images, analyze whether there are blank edge defects in the workpiece; generate corresponding classification signals based on the analysis results. S6: Automatic sorting and unloading: Based on the classification signal, control the unloading mechanism to sort the workpieces and transfer them to the corresponding storage boxes.

10. A method for precision inspection of workpiece blank edges based on machine vision according to claim 9, characterized in that: In step S5, the detected image is processed into a grid, divided into multiple pixel regions, and the contour regions formed by the marker lines in the image are identified; the actual pixel width or area of ​​the contour regions is calculated and compared with a preset standard pixel range. If the actual pixel value does not fall within the standard pixel range, but the deviation from the standard pixel range is within the preset allowable tolerance range, then the workpiece is determined to have a blank edge defect and is a defective part. If the actual pixel value is within the standard pixel range, the workpiece is determined to be free of blank edge defects and is a good product. If the actual pixel value does not fall within the standard pixel range and the deviation from the standard pixel range exceeds the allowable error range, the detection result is deemed invalid and the workpiece is to be inspected.

Images