Panoramic potato defect detection device and method based on combination of virtual imaging and real imaging

Abstract

The invention relates to the technical field of agricultural product detection, and particularly discloses a panoramic potato defect detection device and method based on combination of virtual and real imaging, the panoramic potato defect detection device comprises a rack, the two sides of the rack are fixedly connected with mounting plates, the inner wall of the rack is fixedly connected with a guide rail, and a plurality of station tables circularly move on the guide rail; each station table is rotationally provided with a bearing disc for bearing potato products to move, the mounting plate is provided with two image scanning units, the top of the mounting plate is fixedly connected with a supporting frame, a driving mechanism is arranged in the rack, and the plurality of station tables are controlled to move circularly through the driving mechanism; according to the method, small features and defects of products can be more accurately captured, the accuracy and objectivity of detection results are improved, complete panoramic image data of each potato product can be obtained, the potato products with defects can be more effectively recognized and removed, and the defective rate caused by missing detection is reduced.

Description

Technical Field

[0001] The present invention relates to the technical field of agricultural product detection, and more specifically to a panoramic potato defect detection device and method based on the combination of virtual and real imaging. Background Art

[0002] China is the largest potato producer in the world. Potatoes are one of the four major food crops in modern human society, second only to rice, corn, and wheat. Potatoes contain a large amount of starch, which can provide rich nutritional energy for consumers. With the rapid development of the food industry, the consumption methods of potatoes are becoming more and more diverse, and the demand for mashed potatoes, potato cakes, etc. based on potato powder is increasing.

[0003] Publication No. CN106238342A discloses a panoramic vision potato sorting and defect detection device, including a conveying device, a detection dark box, a sorting mechanism, an infrared sensor module, an image acquisition mechanism, an image processing and analysis module with a convolutional neural network and a support vector machine SVM built-in, a data fusion module with a support vector machine SVM built-in, and a timing module for coordinating the actions of each component; the present invention also discloses a potato sorting and detection method using the above device. The panoramic vision potato sorting and defect detection device and its sorting and detection method of the present invention can complete all-round detection without the need for potatoes to be flipped during the detection process. On the one hand, it avoids unnecessary damage to potatoes, and on the other hand, it avoids the instability in dynamic photo detection, improves the clarity of images, and enhances the accuracy of detection.

[0004] Shape and surface defects are important features of the appearance quality of potatoes. By quantitatively measuring these characteristic indicators, comprehensive detection and grading of external defects, shape, etc. of potatoes are completed. The existing potato quality detection generally relies on manual sensory recognition and judgment. The method of manually detecting potato quality has low efficiency, poor objectivity and accuracy, is difficult to meet the requirements of high-standard grading, is not conducive to realizing large-scale and automated quality detection operations, and the commonly used scanning detection method and conveyor detection method are prone to missed detection and have low detection efficiency for each individual. Therefore, we propose a panoramic potato defect detection device and method based on the combination of virtual and real imaging. Summary of the Invention

[0005] In order to overcome the above-mentioned defects of the prior art, the present invention provides a panoramic potato defect detection device and method based on the combination of virtual and real imaging to solve the problems existing in the above background art.

[0006] The present invention provides the following technical solution: a panoramic potato defect detection device based on the combination of virtual and real imaging, comprising a frame, both sides of the frame are fixedly connected with a mounting plate, the inner wall of the frame is fixedly connected with a guide rail, a plurality of workstations are cyclically moved on the guide rail, each of the workstations is rotatably provided with a carrier plate to carry potato products for movement, two image scanning units are arranged on the mounting plate, the top of the mounting plate is fixedly connected with a support frame, a driving mechanism is arranged in the frame, and the cyclic movement of the plurality of workstations is controlled by the driving mechanism, a clamping and flipping mechanism is arranged on the support frame, and the potato products are clamped and flipped by the clamping and flipping mechanism, and each of the workstations is provided with an intermittent rotating mechanism, and the carrier plate is automatically rotated by the intermittent rotating mechanism, so that the image scanning unit can fully scan the potato products; The intermittent rotation mechanism includes a driven sheave, a positioning shaft, an extension shaft, a spur gear, a driving dial and a bearing seat. The driven sheave is located on the lower side of the work station, the positioning shaft is fixedly connected between the driven sheave and the carrier plate, the bearing seat is fixedly connected to the inner wall of the work station, the extension shaft is rotatably connected to the lower side of the bearing seat, the spur gear and the driving dial are both fixedly connected to the circumferential surface of the extension shaft, and the driving dial is meshed with the driven sheave.

[0007] Furthermore, two racks are fixedly connected to the inner wall of the frame, which are respectively located at the lower side of the corresponding image scanning units, and the spur gear meshes with the racks when passing through them.

[0008] Furthermore, the driving mechanism includes a servo motor, a sprocket, a chain and a plurality of linkage rods, the servo motor is fixedly connected to the side of the frame, two sprockets are provided and are rotatably connected between the inner walls of the frame, the output end of the servo motor is fixedly connected to one of the sprockets, the chain transmission is connected between the two sprockets, and the plurality of linkage rods are respectively fixedly connected between the chain and the corresponding workstations.

[0009] Furthermore, the clamping and flipping mechanism includes a slide seat, a flipping motor, an electric clamp and an electric telescopic rod. The slide seat is slidably connected in the support frame, the flipping motor is fixedly connected to the side of the slide seat, the electric clamp is fixedly connected to the output end of the flipping motor, the electric telescopic rod is fixedly connected to the lower inner wall of the support frame, and the extended end of the electric telescopic rod is fixedly connected to the slide seat.

[0010] Furthermore, guide wheels are installed on both sides of the workstation, and the guide wheels are respectively rollingly connected to the upper and lower sides of the corresponding guide rails.

[0011] Furthermore, the top of the mounting plate is fixedly connected to two protective covers, the two image scanning units are respectively located in different protective covers, and light strips are installed in both protective covers.

[0012] Furthermore, an induction switch is fixedly connected inside the frame, and the induction switch is located at the lower side of the clamping and flipping mechanism. When each workstation moves to the support frame, the servo motor pauses.

[0013] Furthermore, the extended shaft rotates when passing through the rack, and the rotation amplitude of the supporting plate is one circle.

[0014] The method for detecting panoramic potato defects based on the combination of virtual and real imaging includes the following steps: S1. Preparation: Start the servo motor to drive the sprocket to rotate, the sprocket drives the chain to move cyclically, the chain drives the linkage rod to move cyclically, and finally drives several workstations and the carrier plate to move synchronously and cyclically, and turns on the light strip on the protective cover; S2, feeding: conveying the potato products to be tested onto the carrying plate one by one, so that the potato products move along with the carrying plate; S3, first scan: when the carrier plate passes the first rack, the spur gear is used to rotate the active dial, the active dial drives the driven groove wheel to rotate intermittently, the driven groove wheel drives the positioning shaft to rotate intermittently, and finally drives the carrier plate to rotate intermittently, the potato product on the carrier plate rotates intermittently, and the image scanning unit is used to take a photo and scan the potato product during each pause. After passing the rack, the workstation and the carrier plate rotate one circle, that is, the potato product rotates one circle, and image data of the potato product rotating one circle is obtained; S4, flipping: When the workstation passes the induction switch, the servo motor is controlled to pause. At this time, the carrier plate is located at the lower side of the clamping flipping mechanism. The slide is controlled to rise and fall by the electric telescopic rod, and then the potato product is clamped by the electric clamp. Finally, the flipping motor is controlled to drive the electric clamp to flip, so that the potato product is flipped 180 degrees. After the potato product is put back into the carrier plate, the electric clamp returns to its original position. S5, second code scanning: after the potato product is turned over, the servo motor is started again to make the workstation pass through the second rack. According to step S3, image data of the potato product after it is turned over and rotated one circle can be obtained; S6, panoramic analysis: The image data obtained in S3 and S5 are aggregated to obtain complete panoramic image data of each potato product, and then the captured panoramic image data is processed and analyzed to identify defects on the potato surface, such as sprouts, green skin, damage or deformity, etc.; S7. Classification and processing: Based on the recognition results, the potato products are classified and processed, that is, the defective and non-defective potato products are transported to different product lines.

[0015] Technical effects and advantages of the present invention: 1. The present invention is provided with an intermittent rotation mechanism, which is conducive to taking pictures and scanning when the potato products are paused. When the potato products are scanned in the paused state, static images of the potato products can be captured, avoiding blurring and distortion that may occur due to the movement or vibration of the potatoes during dynamic scanning. Static scanning allows for focusing on the details of the products for a longer time, enabling more accurate capture of the minute features and defects of the products, and improving the accuracy and objectivity of the detection results.

[0016] 2. The present invention is provided with a clamping and flipping mechanism, which is conducive to performing a secondary scan after the first scan when the potato products are flipped. After flipping, the bottom of the potato products that was blocked during the first scan can be scanned, and the potato products rotate one full circle again. Combining the image data obtained from the first and second scans, complete panoramic image data of each potato product can be obtained, which can more effectively identify and eliminate defective potato products, reducing the defective rate caused by missed inspections.

[0017] 3. The present invention is provided with a workbench that moves in a cycle, which is conducive to automating the scanning process in conjunction with the cyclic movement of the workbench, improving the overall automation level of the production line, reducing manual intervention and labor intensity, and greatly improving work efficiency.

[0018] 4. The present invention is provided with a protective cover and a light strip, which is conducive to forming a channel-type scanning process, enabling the device to more clearly see the details on the surface of the potatoes, such as flaws, spots, sprouting, damage, etc. This helps to reduce missed inspections and false detections, improving the accuracy of detection. For industrial potato detection, it can significantly improve the detection efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] Figure 1 It is a schematic diagram of the overall structure of the present invention.

[0020] Figure 2 It is a schematic diagram of the structure after the protective cover of the present invention is removed.

[0021] Figure 3 It is a schematic diagram of the internal structure of the present invention.

[0022] Figure 4 It is for the Figure 3 schematic diagram of the structure at location A of the present invention.

[0023] Figure 5 It is a schematic diagram of the intermittent rotation mechanism of the present invention.

[0024] Figure 6 It is a schematic diagram of the clamping and flipping mechanism of the present invention.

[0025] Figure 7 It is a schematic diagram of the sectional structure of the present invention.

[0026] The reference numerals are: 1, frame; 101, mounting plate; 2, guide rail; 3, workbench; 301, guide wheel; 4, carrier plate; 5, image scanning unit; 6, servo motor; 601, sprocket; 602, chain; 603, linkage rod; 7, intermittent rotation mechanism; 701, driven Geneva wheel; 702, positioning rotating shaft; 703, extension shaft; 704, spur gear; 705, driving dial; 706, bearing block; 8, rack; 9, clamping and flipping mechanism; 901, sliding seat; 902, flipping motor; 903, electric clamp; 904, electric telescopic rod; 10, support frame; 11, protective cover; 12, light strip; 13, induction switch. Detailed implementation manners

[0027] Next, the technical solutions in the present invention will be clearly and completely described in conjunction with the accompanying drawings in the present invention. In addition, the forms of each structure described in the following implementation manners are merely examples, and the panoramic potato defect detection device and method based on the combination of virtual and real imaging involved in the present invention are not limited to the structures described in the following implementation manners. All other implementation manners obtained by those of ordinary skill in the art without creative efforts fall within the scope of protection of the present invention.

[0028] Referring to Figures 1 - 7 , the present invention provides a panoramic potato defect detection device based on the combination of virtual and real imaging, including a frame 1. Mounting plates 101 are fixedly connected to both sides of the frame 1. Guide rails 2 are fixedly connected to the inner wall of the frame 1. A number of workbenches 3 are movably circulated on the guide rails 2. A carrier plate 4 is rotatably arranged on each workbench 3 to carry potato products for movement. Two image scanning units 5 are arranged on the mounting plates 101. A support frame 10 is fixedly connected to the top of the mounting plates 101. A driving mechanism is arranged inside the frame 1 to control the cyclic movement of a number of workbenches 3. A clamping and flipping mechanism 9 is arranged on the support frame 10 to clamp and flip potato products. An intermittent rotation mechanism 7 is arranged on each workbench 3 to control the automatic rotation of the carrier plate 4, so that the image scanning unit 5 can comprehensively scan potato products.

[0029] The intermittent rotation mechanism 7 includes a driven Geneva wheel 701, a positioning rotating shaft 702, an extension shaft 703, a spur gear 704, a driving dial 705 and a bearing block 706. The driven Geneva wheel 701 is located below the workbench 3. The positioning rotating shaft 702 is fixedly connected between the driven Geneva wheel 701 and the carrier plate 4. The bearing block 706 is fixedly connected to the inner wall of the workbench 3. The extension shaft 703 is rotatably connected to the lower side of the bearing block 706. The spur gear 704 and the driving dial 705 are both fixedly connected to the circumferential surface of the extension shaft 703. The driving dial 705 meshes with the driven Geneva wheel 701.

[0030] In this embodiment, it should be specifically noted that: the guide rail 2 is used to restrict the cyclic movement of several workstations 3. The several workstations 3 are evenly distributed, and the distance between adjacent workstations 3 is the same. The carrier plate 4 is used to carry the potato products to be detected. During use, the servo motor 6 is used to control the rotation of the sprocket 601. Under the driving action of the chain 602, several linkage rods 603 are driven to move synchronously, and finally several workstations 3 and the carrier plate 4 are driven to move synchronously. The image scanning unit 5 is used to scan the potato products and is composed of multiple scanners. When each workstation 3 passes by the rack 8, since the spur gear 704 meshes with the rack 8, the spur gear 704 rotates self-driven when passing by the rack 8, and then drives the extension shaft 703 and the driving dial 705 to rotate synchronously. The driving dial 705 drives the driven Geneva wheel 701 to rotate intermittently, and the driven Geneva wheel 701 drives the positioning rotating shaft 702 to rotate intermittently. Finally, the carrier plate 4 is driven to rotate intermittently, and the potato products located in the carrier plate 4 rotate intermittently. After passing by the rack 8, the workstation 3 and the carrier plate 4 rotate one week, that is, the potato products rotate one week, and the image data of the potato products rotating one week is obtained. This solution is set to take pictures and scans when the potato products are paused. When the potato products are scanned in the paused state, static images of the potato products can be captured, avoiding blurring and distortion that may occur due to the movement or vibration of the potatoes during dynamic scanning. Static scanning allows for a longer time to focus on the details of the product, so that the minute features and defects of the product can be captured more accurately. The clear static images provide a more accurate basis for data collection, helping to generate more accurate product information and quality control reports. Moreover, since pausing the scan can provide clearer images, the number of times of re-scanning due to image blurring or distortion is reduced. Especially in the production line, pausing the scan can be combined with the automated process to achieve more efficient detection and quality control. The workstations 3 that can move cyclically can adapt to the current industrial equipment, greatly improving the work efficiency.

[0031] The main difference between this embodiment and the prior art lies in the use of a secondary scanning detection method in this embodiment. Specifically, after the first scan, the carrier plate 4 will move to the lower side of the clamping and flipping mechanism 9 and pause. The electric telescopic rod 904 is used to control the lifting of the sliding seat 901, and then the electric clamp 903 is used to clamp the potato product. Finally, the flipping motor 902 is controlled to drive the electric clamp 903 to flip, so that the potato product is flipped 180 degrees. After the potato product is flipped, a secondary scan is performed. After flipping, the bottom of the potato product that was blocked during the first scan can be scanned, and the potato product rotates one full circle again. By combining the image data obtained from the first and second scans, the complete panoramic image data of each potato product can be obtained. Then, the panoramic image data captured is processed and analyzed using intelligent algorithms to identify defects on the potato surface, such as sprouting, green skin, damage, or deformity, etc. A virtual image forming model of the potato product is constructed. The intelligent algorithm can use image segmentation algorithms (such as U-Net, Mask R-CNN) to perform pixel-level segmentation of potato surface defects, accurately identify the shape and size of the defects, which is suitable for refined analysis of complex defects. It can also use object detection algorithms (such as YOLO, Faster R-CNN) to quickly locate the position and range of potato surface defects, which is suitable for real-time detection scenarios and improves the detection efficiency. Through flipping and scanning, defective potato products can be more effectively identified and removed, reducing the defective rate caused by missed inspections and reducing the additional costs caused by rework or scrapping. Moreover, the flipping and scanning process of this solution and the cyclic movement of the workstation 3 achieve an automated process, improving the overall automation level of the production line and reducing manual intervention and labor intensity.

[0032] The above structure is the main structure of this embodiment, which solves the problem of low detection efficiency and accuracy of potato product defects. The cyclic movement of the workstation 3 on the guide rail 2 is an existing structure. The specific structure and connection method of the electric clamp 903 for controlling the flipping of the potato product are not specifically described in this embodiment. In addition, the image scanning unit 5 scanning the potato product also belongs to the prior art. Therefore, this application does not make a detailed limitation.

[0033] Refer to Figure 3 、 Figure 4 and Figure 5 As shown in

[0034] In this embodiment, it should be specifically noted that the height of the rack 8 is the same as that of the spur gear 704. When the workbench 3 passes by the rack 8, the spur gear 704 will mesh with the rack 8. Then, as the workbench 3 continues to move, the spur gear 704 rotates on its own axis, driving the driving dial 705 to rotate, and finally driving the bearing plate 4 to rotate intermittently. The rotation amplitude of the bearing plate 4 is one full circle to ensure that the potato products in the bearing plate 4 rotate one full circle, obtaining complete panoramic image data, and clearly observing defects such as cracks, lesions, insect eyes, and mechanical damages on the potato surface.

[0035] Referring to Figure 3 , the driving mechanism includes a servo motor 6, a sprocket 601, a chain 602, and a number of linkage rods 603. The servo motor 6 is fixedly connected to the side of the frame 1. There are two sprockets 601, both of which are rotatably connected between the inner walls of the frame 1. The output end of the servo motor 6 is fixedly connected to one of the sprockets 601. The chain 602 is drivingly connected between the two sprockets 601. A number of linkage rods 603 are respectively fixedly connected between the chain 602 and the corresponding workbench 3.

[0036] In this embodiment, it should be specifically noted that the servo motor 6 is a commonly used industrial motor. By driving one of the sprockets 601 through the output end of the servo motor 6, the chain 602 is then driven to move in a cycle. Under the connection of the linkage rods 603, a number of workbenches 3 and bearing plates 4 are driven to move synchronously in a cycle. Using the linkage rods 603 as the transmission medium can accurately and stably drive the workstations to move in a cycle, thereby realizing the automated operation of the production line. This greatly improves production efficiency, reduces manual intervention, and reduces the risk of human errors. It has excellent load-bearing capacity and wear resistance, can adapt to various heavy-load and high-speed production environments, and ensures the stable operation of the production line.

[0037] Referring to Figure 4 , the clamping and flipping mechanism 9 includes a sliding seat 901, a flipping motor 902, an electric clamp 903, and an electric telescopic rod 904. The sliding seat 901 is slidably connected within the support frame 10. The flipping motor 902 is fixedly connected to the side of the sliding seat 901. The electric clamp 903 is fixedly connected to the output end of the flipping motor 902. The electric telescopic rod 904 is fixedly connected to the lower inner wall of the support frame 10, and the extending end of the electric telescopic rod 904 is fixedly connected to the sliding seat 901.

[0038] In this embodiment, it should be specifically noted that: under the restriction of the support frame 10, the sliding seat 901 can only move linearly. The extending end of the electric telescopic rod 904 can push the sliding seat 901 to lift, so as to adjust the height of the electric fixture 903 to adapt to potato products of different heights. The electric fixture 903 is an existing parallel clamping tool. After the sliding seat 901 moves to a suitable height, the electric fixture 903 can be controlled to clamp the potato product, and then the flipping motor 902 is controlled to flip, driving the potato product to flip 180 degrees. Then the potato product is placed back in place. After the potato product is flipped, a secondary scan is performed. After flipping, the bottom of the potato product that was blocked during the first scan can be scanned, and complete panoramic image data of each potato product can be obtained.

[0039] Refer to Figure 3 and Figure 5 , guide wheels 301 are installed on both sides of the workbench 3, and the guide wheels 301 are respectively connected to the upper and lower sides of the corresponding guide rails 2 in a rolling manner.

[0040] In this embodiment, it should be specifically noted that: the guide rails 2 are symmetrically arranged on the inner wall of the frame 1, and the guide wheels 301 roll on the upper and lower sides of the guide rails 2 to ensure that the workbench 3 can move stably without deviation. The structure is stable and reasonable, and the service life is long.

[0041] Refer to Figure 4 , the top of the mounting plate 101 is fixedly connected to two protective covers 11, and the two image scanning units 5 are respectively located in different protective covers 11. Light strips 12 are installed in both protective covers 11.

[0042] In this embodiment, it should be specifically noted that: the two protective covers 11 are both located on the upper side of the frame 1. Each carrier plate 4 will pass through the inside of the protective cover 11. Using the protective cover 11 to form a channel-type scanning process, the detection of a large number of potatoes can be completed in a short time, which can significantly improve the detection efficiency, reduce the influence of other factors, reduce the time and cost of manual detection. The designed light strip 12 increases the illumination, which can improve the light reflectivity of the potato surface, enabling the device to see the details of the potato surface more clearly, such as flaws, spots, sprouting, damage, etc. This helps to reduce missed detections and false detections and improve the accuracy of detection. For industrial potato detection, the detection efficiency can be significantly improved.

[0043] Refer to Figure 4 , an induction switch 13 is fixedly connected inside the frame 1. The induction switch 13 is located below the clamping and flipping mechanism 9. When each workbench 3 moves to the support frame 10, the servo motor 6 pauses.

[0044] In this embodiment, it should be specifically noted that the induction switch 13 is a prior art. When the workbench 3 passes by the induction switch 13, the linkage rod 603 will trigger the induction switch 13, causing the servo motor 6 to pause. After the clamping and flipping mechanism 9 flips the potato product, the servo motor 6 resumes operation. The specific structure will not be elaborated in detail.

[0045] The working principle of the present invention: The main problems solved in this embodiment are as follows: By using an automated scanning method to obtain complete panoramic image data of potato products, it solves the problem of poor objectivity and accuracy in existing detection methods. By using an automated individual detection method, the work efficiency is high and there is no missed detection, solving the problem of easy missed detection and low individual detection efficiency in existing conveyor-type detection methods.

[0046] A detection method for panoramic potato defects based on the combination of virtual and real imaging includes the following steps: S1. Preparation: Start the servo motor 6 to drive the sprocket 601 to rotate. The sprocket 601 drives the chain 602 to move in a cycle. The chain 602 drives the linkage rod 603 to move in a cycle, and finally drives a number of workbenches 3 and carrier plates 4 to move synchronously in a cycle, and turn on the light strip 12 on the protective cover 11; S2. Feeding: Feed the potato products to be detected one by one onto the carrier plate 4, so that the potato products move with the carrier plate 4; S3. First scan: When the carrier plate 4 passes the first rack 8, use the rotation of the spur gear 704 to drive the active dial 705 to rotate. The active dial 705 drives the driven Geneva wheel 701 to rotate intermittently. The driven Geneva wheel 701 drives the positioning rotating shaft 702 to rotate intermittently, and finally drives the carrier plate 4 to rotate intermittently. The potato products on the carrier plate 4 rotate intermittently. Each time it pauses, the image scanning unit 5 takes a photo scan of the potato products. After passing the rack 8, the workbench 3 and the carrier plate 4 rotate one week, that is, the potato products rotate one week, and the image data of the potato products rotating one week is obtained; S4. Flipping: When the workbench 3 passes by the induction switch 13, control the servo motor 6 to pause. At this time, the carrier plate 4 is located below the clamping and flipping mechanism 9. Control the lifting of the sliding seat 901 through the electric telescopic rod 904, then clamp the potato product through the electric clamp 903, and finally control the flipping motor 902 to drive the electric clamp 903 to flip, so that the potato product flips 180 degrees. After putting the potato product back into the carrier plate 4, the electric clamp 903 returns to its original position; S5. Second scanning: After the potato product is flipped, start the servo motor 6 again, so that the workbench 3 passes the second rack 8. According to step S3, the image data of the potato product after flipping and rotating one week can be obtained; S6. Panoramic analysis: Summarize the image data obtained in S3 and S5 to obtain the complete panoramic image data of each potato product. Then, process and analyze the captured panoramic image data to identify defects on the potato surface, such as sprouting, green skin, breakage, or deformity, etc. S7. Classification processing: According to the recognition results, classify and process the potato products, that is, convey the potato products with or without defects to different production lines.

[0047] The above are only the preferred embodiments of the present invention and are not intended to limit the present invention. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principle of the present invention shall be included within the protection scope of the present invention.

Claims

1. A panoramic potato defect detection device based on a combination of virtual and real imaging, comprising a frame (1), characterized in that: Both sides of the frame (1) are fixedly connected with mounting plates (101); the inner wall of the frame (1) is fixedly connected with a guide rail (2); a plurality of workstations (3) are cyclically movable on the guide rail (2); each workstation (3) is rotatably provided with a carrier plate (4) for carrying potato products for movement; two image scanning units (5) are provided on the mounting plate (101); the top of the mounting plate (101) is fixedly connected with a support frame (10); a driving mechanism is provided in the frame (1) for controlling the cyclic movement of the plurality of workstations (3); a clamping and flipping mechanism (9) is provided on the support frame (10); the potato products are clamped and flipped by the clamping and flipping mechanism (9); each workstation (3) is provided with an intermittent rotating mechanism (7); the carrier plate (4) is automatically rotated by the intermittent rotating mechanism (7) so that the image scanning unit (5) can fully scan the potato products; The intermittent rotation mechanism (7) comprises a driven sheave (701), a positioning shaft (702), an extension shaft (703), a spur gear (704), a driving dial (705) and a bearing seat (706); the driven sheave (701) is located at the lower side of the work station (3); the positioning shaft (702) is fixedly connected between the driven sheave (701) and the carrier plate (4); the bearing seat (706) is fixedly connected to the inner wall of the work station (3); the extension shaft (703) is rotatably connected to the lower side of the bearing seat (706); the spur gear (704) and the driving dial (705) are both fixedly connected to the circumferential surface of the extension shaft (703); and the driving dial (705) is meshed with the driven sheave (701).

2. The panoramic potato defect detection device based on the combination of virtual and real imaging according to claim 1 is characterized in that: Two racks (8) are fixedly connected to the inner wall of the frame (1), which are respectively located at the lower side of the corresponding image scanning unit (5), and the spur gear (704) meshes with the racks (8) when passing through them.

3. The panoramic potato defect detection device based on the combination of virtual and real imaging according to claim 1 is characterized in that: The driving mechanism comprises a servo motor (6), a sprocket (601), a chain (602) and a plurality of linkage rods (603); the servo motor (6) is fixedly connected to a side of a frame (1); two sprockets (601) are provided and are rotatably connected between inner walls of the frame (1); an output end of the servo motor (6) is fixedly connected to one of the sprockets (601); the chain (602) is transmission-connected between the two sprockets (601); and the plurality of linkage rods (603) are respectively fixedly connected between the chain (602) and a corresponding workstation (3).

4. The panoramic potato defect detection device based on the combination of virtual and real imaging according to claim 1 is characterized in that: The clamping and flipping mechanism (9) comprises a slide seat (901), a flipping motor (902), an electric clamp (903) and an electric telescopic rod (904); the slide seat (901) is slidably connected to the support frame (10); the flipping motor (902) is fixedly connected to the side of the slide seat (901); the electric clamp (903) is fixedly connected to the output end of the flipping motor (902); the electric telescopic rod (904) is fixedly connected to the lower inner wall of the support frame (10); and the extended end of the electric telescopic rod (904) is fixedly connected to the slide seat (901).

5. The panoramic potato defect detection device based on the combination of virtual and real imaging according to claim 1 is characterized in that: Guide wheels (301) are installed on both sides of the workstation (3), and the guide wheels (301) are respectively connected in a rolling manner to the upper and lower sides of the corresponding guide rails (2).

6. The panoramic potato defect detection device based on the combination of virtual and real imaging according to claim 1 is characterized in that: The top of the mounting plate (101) is fixedly connected to two protective covers (11); the two image scanning units (5) are respectively located in different protective covers (11); and a light strip (12) is installed in each of the two protective covers (11).

7. The panoramic potato defect detection device based on the combination of virtual and real imaging according to claim 4 is characterized in that: An inductive switch (13) is fixedly connected inside the frame (1), and the inductive switch (13) is located at the lower side of the clamping and flipping mechanism (9). When each workstation (3) moves to the support frame (10), the servo motor (6) is paused.

8. The panoramic potato defect detection device based on the combination of virtual and real imaging according to claim 2 is characterized in that: The extended shaft (703) generates self-rotation when passing through the rack (8), and the bearing plate (4) rotates once.

9. A method for detecting panoramic potato defects based on a combination of virtual and real imaging, applied to a panoramic potato defect detection device based on a combination of virtual and real imaging according to claim 8, characterized in that: The steps include: S1, preparation: start the servo motor (6) to drive the sprocket (601) to rotate, the sprocket (601) drives the chain (602) to move cyclically, the chain (602) drives the linkage rod (603) to move cyclically, and finally drives a plurality of workstations (3) and the carrier plate (4) to move synchronously and cyclically, and opens the light strip (12) on the protective cover (11); S2, feeding: conveying the potato products to be tested onto the carrying plate (4) one by one, so that the potato products move along with the carrying plate (4); S3, first scanning: when the carrier plate (4) passes the first rack (8), the spur gear (704) is used to rotate to drive the active dial (705), the active dial (705) drives the driven groove wheel (701) to rotate intermittently, the driven groove wheel (701) drives the positioning shaft (702) to rotate intermittently, and finally drives the carrier plate (4) to rotate intermittently, the potato product on the carrier plate (4) rotates intermittently, and the image scanning unit (5) is used to take a photo and scan the potato product during each pause. After passing the rack (8), the workstation (3) and the carrier plate (4) rotate one circle, that is, the potato product rotates one circle, and image data of the potato product rotating one circle is obtained; S4, flipping: when the workstation (3) passes the induction switch (13), the servo motor (6) is controlled to pause, and at this time the carrier plate (4) is located at the lower side of the clamping flipping mechanism (9), the slide seat (901) is controlled to rise and fall by the electric telescopic rod (904), and then the potato product is clamped by the electric clamp (903), and finally the flipping motor (902) is controlled to drive the electric clamp (903) to flip, so that the potato product is flipped 180 degrees, and after the potato product is placed back into the carrier plate (4), the electric clamp (903) returns to its original position; S5, second code scanning: after the potato product is turned over, the servo motor (6) is started again to make the workstation (3) pass through the second rack (8). According to step S3, image data of the potato product after it is turned over and rotated one circle can be obtained; S6, panoramic analysis: The image data obtained in S3 and S5 are aggregated to obtain complete panoramic image data of each potato product, and then the captured panoramic image data is processed and analyzed to identify defects on the potato surface, such as sprouts, green skin, damage or deformity, etc.; S7. Classification and processing: Based on the recognition results, the potato products are classified and processed, that is, the defective and non-defective potato products are transported to different product lines.

Images