A potato diffuse transmission hyperspectral image acquisition device and method

Abstract

The present invention discloses a potato diffuse transmission hyperspectral image acquisition device and method, and relates to the technical field of potato quality detection. A potato diffuse transmission hyperspectral image acquisition device comprises a bottom plate, and support legs are welded around the bottom outer wall of the bottom plate, and a collection box is installed at one end of the top outer wall of the bottom plate. In the present invention, when working, the potato slices to be detected are laid flat on the light-transmitting plate, and the light source lamp is turned on at the same time, and the collection tube is inserted into the collection box. As the cutting ring contacts the potato, the irregular parts of the potato edge are cut off, and the remaining parts are completely merged with the collection tube. The potato detection area is limited by the collection tube, and the light transmission problem caused by the irregular edge of the potato is avoided. The spectral image acquisition is more accurate and without interference. At the same time, the collection box blocks external light to prevent the light source in the outer wall environment from affecting the collection of the spectral image, and the spectral image acquisition is more accurate.

Description

Technical Field

[0001] The invention relates to the technical field of potato quality detection, and in particular to a potato diffuse transmission hyperspectral image acquisition device and method. Background Art

[0002] At present, potato quality detection is an urgent problem to be solved in the fields of potato production, storage, transportation and processing. Conventional biological and chemical detection is time-consuming and labor-intensive. Research in recent years has shown that the use of hyperspectral imaging technology for non-destructive detection of agricultural product quality is an important development trend. Introducing hyperspectral imaging technology into potato quality detection can improve the detection efficiency and accuracy of potato quality and eliminate errors caused by human subjective factors.

[0003] The shortcomings of the existing potato diffuse transmission hyperspectral image acquisition device are:

[0004] When the existing potato diffuse transmission hyperspectral image acquisition device is used for acquisition, due to the uneven potato slices, the edge of the potato is easily disturbed by light during detection, thus affecting image acquisition;

[0005] When the existing potato diffuse transmission hyperspectral image acquisition device is collecting images, external light sources are likely to interfere with the image, resulting in errors in image acquisition and insufficient precision.

[0006] When the existing potato diffuse transmission hyperspectral image acquisition device is performing spectral image analysis, it is greatly affected by the external environment, and the analysis results are prone to deviation and have certain errors.

[0007] Therefore, a potato diffuse transmission hyperspectral image acquisition device and method are proposed to solve the above problems. Summary of the invention

[0008] Based on the technical problems existing in the background technology, the present invention proposes a potato diffuse transmission hyperspectral image acquisition device and method.

[0009] The present invention provides a potato diffuse transmission hyperspectral image acquisition device, comprising a base plate, wherein support legs are welded around the bottom outer wall of the base plate, a collection box is installed at one end of the top outer wall of the base plate, the collection box is made of opaque material, the collection box is adapted to the base plate, a control component is installed at one end of the base plate close to the collection box, the control component is connected to a collection component, the collection component is adapted to the control component, a light-transmitting plate is slidably installed on the inner wall of the collection box, the light-transmitting plate is adapted to the collection box, light source lamps are arranged on the inner wall of the bottom of the collection box at equal distances, the light-transmitting plate is adapted to the collection component, the collection component comprises a mounting plate, collection tubes are installed at the bottom of the mounting plate at equal distances, a hyperspectral imager is installed on the top inner wall of the collection tube, the hyperspectral imager is adapted to the collection tube, a cutting ring is installed on the bottom outer wall of the collection tube, and the diameter of the cutting ring is the same as the diameter of the collection tube.

[0010] Preferably, mounting blocks are slidably installed on the inner walls on both sides of the collection box, the light-transmitting plate is fixedly connected to the mounting blocks, the light-transmitting plate is adapted to the mounting blocks, adjustment grooves are provided on the inner walls on both sides of the collection box, the mounting blocks are slidably connected to the inner walls of the adjustment grooves, a threaded rod is rotatably installed on the inner walls of the adjustment grooves, the length of the threaded rod is the same as the length of the adjustment groove, and the mounting block is rotatably connected to the threaded rod through threads.

[0011] Preferably, driving grooves are provided at the bottom of the inner walls on both sides of the collection box, and a servo motor is installed on the inner wall of the driving groove, and the output shaft of the servo motor is connected to the threaded rod.

[0012] Preferably, fixing blocks are provided on the outer walls on both sides of the light-transmitting plate, and the outer walls on both adjacent sides of the mounting block are provided with fixing grooves distributed at equal distances, the fixing blocks are adapted to the fixing grooves, and fixing holes are provided on the outer walls of the fixing blocks and the inner walls at the bottom of the fixing grooves, and fixing bolts are threadedly installed on the inner walls of the fixing holes, and the length of the fixing bolts is greater than the thickness of the fixing blocks.

[0013] Preferably, the control component includes a mounting rod welded to one end of the top outer wall of the base plate, a mounting bracket is welded to the top of the mounting rod, the mounting bracket is a "T"-shaped structure, a mounting hole is opened on the top outer wall of the mounting bracket, an electric telescopic rod is installed on the inner wall of the mounting hole, and the end of the electric telescopic rod is connected to the mounting plate.

[0014] Preferably, a box cover is slidably installed on the top of the collection box, a collection slot is provided on the top of the box cover, the collection slot is adapted to the mounting plate, a control block is provided at one end of the top outer wall of the box cover, slide slots are provided at both ends of the top outer wall of the collection box, sliders are welded at both ends of the bottom outer wall of the box cover, and the sliders are slidably connected to the inner walls of the slide slots.

[0015] Preferably, a collection computer is installed at one end of the top outer wall of the base plate, the collection computer is connected to a data processor via a wireless network, the data processor is connected to an image analyzer via a wireless network, and the image analyzer is connected to a hyperspectral imager via a wireless network.

[0016] Preferably, a control panel is installed on the top outer wall of the base plate close to the acquisition computer, and control buttons are arranged on the outer wall of the control panel, and the control buttons are connected to the electrical components through wires.

[0017] Preferably, a mounting groove is provided on the top outer wall of the bottom plate, the mounting groove is adapted to the collection box, and the depth of the mounting groove is less than the height of the collection box.

[0018] A potato diffuse transmission hyperspectral image acquisition method comprises the following steps:

[0019] S1: Image acquisition: Before the test, select healthy potatoes and diseased potatoes for slicing, classify the sliced ​​potatoes for image acquisition, slice the potatoes to be tested, open the box cover after slicing, place the potatoes on the light-transmitting plate, and then drive the mounting plate to descend so that the collection tube contacts the potatoes. As the collection tube continues to descend, the cutting ring cuts the potatoes, leaving only potatoes with the same area as the collection tube. At this time, the light source lamp at the bottom of the collection box starts to work to supplement the light source, and the light source is transmitted to the potatoes through the light-transmitting plate. At the same time, the hyperspectral imager collects images of the potatoes. The collected spectral images of the potatoes are processed by the analyzer and the processor, and the corresponding images and data intervals are obtained on the collection computer. In subsequent tests, if the values ​​of the spectral images of the potato slices are all within the value interval, it means that the potatoes are healthy;

[0020] S2: Model establishment: After image acquisition, the anthocyanin, starch and protein contents of the potatoes used for image acquisition are measured, and the data range of the anthocyanin, starch and protein contents of normal potatoes is determined first, and then the data range of the anthocyanin, starch and protein contents of potatoes infected with ring rot is determined, and the data of healthy potatoes and infected potatoes are compared to determine the impact of potato infection with ring rot on the anthocyanin, starch and protein contents, and the spectral image of healthy potatoes and the anthocyanin, starch and protein data of healthy potatoes are classified, and the spectral image of infected potatoes and the anthocyanin, starch and protein data of infected potatoes are classified, and the regression model of potatoes infected with ring rot and not infected with ring rot and the classification model of the infection degree of ring rot are established based on the anthocyanin, starch and protein data of the spectral image, and the comparison of the spectral image with the anthocyanin, starch and protein data can determine whether the potato is infected, and the obtained model is written into the analysis module of the acquisition computer;

[0021] S3: Model verification: After the model is established, potato slices are taken for image acquisition. The acquired images are automatically input into the regression model and the classification model for analysis and processing. The acquisition computer displays the analysis results to determine whether the potato suffers from ring rot. Then, the potato used for verification is stained and other detection processes are performed. The detection structure is compared with the model verification result to observe whether the two results are consistent, and the accuracy of the model is reversely verified. If the results are consistent, it means that the model is successfully established. If the results are inconsistent, adjustments are continued;

[0022] S4: Application of the model: After the model is verified to be able to operate and work normally, potatoes can be directly inspected. When inspecting potatoes, open the box cover and place the potatoes on the light-transmitting plate. Then the electric telescopic rod drives the mounting plate to descend so that the collection tube contacts the potatoes. As the collection tube continues to descend, the cutting ring cuts the potatoes, leaving only potatoes with the same area as the collection tube. At this time, the light source lamp at the bottom of the collection box starts to work to supplement the light source. The light source is transmitted onto the potatoes through the light-transmitting plate. At the same time, the hyperspectral imager collects images of the potatoes. After the collected images are processed by the analysis module, the corresponding images and detection results are obtained on the collection computer.

[0023] The beneficial effects of the present invention are:

[0024] 1. When working, lay the potato slices to be inspected on the light-transmitting plate, turn on the light source lamp, and insert the mounting plate into the collection box under the action of the control component. As the cutting ring contacts the potato, the irregular parts of the potato edge are cut off, and the remaining parts are completely merged with the collection tube. Compared with the traditional image acquisition device, the potato detection area is limited by the collection tube, which avoids the light transmission problem caused by the irregular edge of the potato. The spectral image acquisition is more accurate and has no interference. At the same time, the collection box blocks the external light to prevent the light source in the outer wall environment from affecting the collection of the spectral image, and the spectral image acquisition is more accurate.

[0025] 2. When collecting spectral images of potatoes, the servo motor drives the threaded rod to rotate, which can drive the mounting block to move up and down, thereby adjusting the distance between the light-transmitting plate and the light source lamp, thereby adjusting the light source intensity according to the thickness of the potato slices, making the operation more convenient.

[0026] 3. The present invention establishes a relevant model for potato detection by combining potato detection and processing with hyperspectral imaging. The potato spectral image information is imported into the model for analysis, and the useful information is finally extracted in the shortest time through layer-by-layer feature extraction. The deep model can automatically learn complex information and modify the model depth and parameters according to the difficulty of the features, so that potatoes can be detected quickly and accurately, and the multi-component contents of potato anthocyanins, starch, and protein can be simultaneously determined, and the severity of ring rot can be determined. BRIEF DESCRIPTION OF THE DRAWINGS

[0027] Figure 1 This is a schematic diagram of the main appearance structure of a potato diffuse transmission hyperspectral image acquisition device and method proposed by the present invention;

[0028] Figure 2 A schematic diagram of the structure of a hyperspectral imager for a potato diffuse transmission hyperspectral image acquisition device and method proposed in the present invention;

[0029] Figure 3 This is a schematic diagram of the working state structure of the acquisition component of a potato diffuse transmission hyperspectral image acquisition device and method proposed by the present invention;

[0030] Figure 4 This is a schematic diagram of the installation slot structure of a potato diffuse transmission hyperspectral image acquisition device and method proposed by the present invention;

[0031] Figure 5 A schematic diagram of the cross-sectional structure of a collection box of a potato diffuse transmission hyperspectral image collection device and method proposed by the present invention;

[0032] Figure 6 This is a schematic diagram of the light-transmitting plate structure of a potato diffuse transmission hyperspectral image acquisition device and method proposed by the present invention;

[0033] Figure 7 This is a schematic diagram of the threaded rod structure of a potato diffuse transmission hyperspectral image acquisition device and method proposed by the present invention.

[0034] In the figure: 1. bottom plate; 2. collection box; 3. light-transmitting plate; 4. light source lamp; 5. mounting plate; 6. collection tube; 7. hyperspectral imager; 8. cutting ring; 9. mounting block; 10. adjustment slot; 11. threaded rod; 12. servo motor; 13. fixing block; 14. fixing slot; 15. fixing hole; 16. fixing bolt; 17. mounting rod; 18. mounting bracket; 19. electric telescopic rod; 20. box cover; 21. collection slot; 22. control block; 23. collection computer; 24. control panel; 25. control button; 26. mounting slot. DETAILED DESCRIPTION

[0035] The technical solutions in the embodiments of the present invention will be described clearly and completely below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all the embodiments.

[0036] Example, see Figure 1-7 A potato diffuse transmission hyperspectral image acquisition device and method, comprising a bottom plate 1, the bottom outer wall of the bottom plate 1 is welded with support legs all around, a collection box 2 is installed at one end of the top outer wall of the bottom plate 1, the collection box 2 is made of opaque material, the collection box 2 is adapted to the bottom plate 1, a control component is installed at one end of the bottom plate 1 close to the collection box 2, the control component is connected to the collection component, the collection component is adapted to the control component, a light-transmitting plate 3 is slidably installed on the inner wall of the collection box 2, the light-transmitting plate 3 is adapted to the collection box 2, the inner wall of the bottom of the collection box 2 is provided with light source lamps 4 distributed at equal distances, the light-transmitting plate 3 is adapted to the collection component, the collection component comprises a mounting plate 5, and the bottom of the mounting plate 5 is installed with collection tubes 6 distributed at equal distances A hyperspectral imager 7 is installed on the inner wall of the top of the collection tube 6, and the hyperspectral imager 7 is adapted to the collection tube 6. A cutting ring 8 is installed on the outer wall of the bottom of the collection tube 6, and the diameter of the cutting ring 8 is the same as the diameter of the collection tube 6. A box cover 20 is slidably installed on the top of the collection box 2, and a collection groove 21 is opened on the top of the box cover 20. The collection groove 21 is adapted to the mounting plate 5. A control block 22 is arranged at one end of the top outer wall of the box cover 20. Slide grooves are opened at both ends of the top outer wall of the collection box 2, and sliders are welded at both ends of the bottom outer wall of the box cover 20. The slider is slidably connected to the inner wall of the slide groove, and an installation groove 26 is opened on the top outer wall of the bottom plate 1. The installation groove 26 is adapted to the collection box 2, and the depth of the installation groove 26 is less than the height of the collection box 2.

[0037] By means of the above mechanism: when working, the potatoes to be inspected are sliced, and then the box cover 20 is opened, and the potatoes to be inspected are laid flat on the light-transmitting plate 3 according to the corresponding position of the collection tube 6, and the light source lamp 4 is turned on at the same time. At this time, the mounting plate 5 moves downward under the action of the control component and is inserted into the collection box 2. As the cutting ring 8 contacts the potato, the irregular parts of the potato edge are cut off, and the remaining parts are completely merged with the collection tube 6. Compared with the traditional image acquisition device, the potato detection area is limited by the collection tube 6, which avoids the light transmission problem caused by the irregular edge of the potato, and the spectral image acquisition is more accurate and without interference. At the same time, the collection box 2 blocks the external light to prevent the light source in the external wall environment from affecting the acquisition of the spectral image, and the spectral image acquisition is more accurate.

[0038] As a preferred implementation of this embodiment, mounting blocks 9 are slidably installed on the inner walls of both sides of the collection box 2, the light-transmitting plate 3 is fixedly connected to the mounting blocks 9, the light-transmitting plate 3 is adapted to the mounting blocks 9, and the inner walls of both sides of the collection box 2 are provided with adjustment grooves 10, the mounting blocks 9 are slidably connected to the inner walls of the adjustment grooves 10, and the inner walls of the adjustment grooves 10 are rotatably installed with threaded rods 11, the length of the threaded rods 11 is the same as the length of the adjustment grooves 10, the mounting blocks 9 are rotatably connected to the threaded rods 11 through threads, and the bottom of the inner walls of both sides of the collection box 2 are provided with drive grooves, and the inner walls of the drive grooves are installed with servo motors 12, and the output shafts of the servo motors 12 are connected to the threaded rods 11, and the light-transmitting plates 3 are on both sides. The outer walls are provided with fixing blocks 13, and the outer walls on both sides of the adjacent mounting blocks 9 are provided with fixing grooves 14 distributed at equal distances. The fixing blocks 13 are adapted to the fixing grooves 14, and the outer walls of the fixing blocks 13 and the inner walls at the bottom of the fixing grooves 14 are provided with fixing holes 15. The inner walls of the fixing holes 15 are threadedly installed with fixing bolts 16, and the length of the fixing bolts 16 is greater than the thickness of the fixing blocks 13. When the spectral image of potatoes is collected, the threaded rod 11 is driven to rotate by the servo motor 12 to drive the mounting blocks 9 to move up and down, thereby adjusting the distance between the light-transmitting plate 3 and the light source lamp 4, thereby adjusting the light source intensity according to the thickness of the potato slices, and the operation is more convenient.

[0039] As a preferred implementation of this embodiment, the control component includes a mounting rod 17 welded to one end of the top outer wall of the base plate 1, a mounting frame 18 is welded to the top of the mounting rod 17, the mounting frame 18 is a "T"-shaped structure, a mounting hole is opened on the top outer wall of the mounting frame 18, an electric telescopic rod 19 is installed on the inner wall of the mounting hole, and the end of the electric telescopic rod 19 is connected to the mounting plate 5. When potatoes are tested, the electric telescopic rod 19 drives the mounting plate 5 to rise and fall, thereby controlling the collection component, and the operation is more convenient.

[0040] As a preferred implementation manner of this embodiment, a collection computer 23 is installed at one end of the top outer wall of the base plate 1, the collection computer 23 is connected to a data processor via a wireless network, the data processor is connected to an image analyzer via a wireless network, the image analyzer is connected to the hyperspectral imager 7 via a wireless network, and a control panel 24 is installed on the top outer wall of the base plate 1 close to the collection computer 23, and a control button 25 is arranged on the outer wall of the control panel 24, and the control button 25 is connected to the electrical components via wires.

[0041] A potato diffuse transmission hyperspectral image acquisition method comprises the following steps:

[0042] S1: Image acquisition: Before the test, select healthy potatoes and diseased potatoes for slicing, classify the sliced ​​potatoes for image acquisition, slice the potatoes to be tested, open the box cover 20 after slicing, place the potatoes on the light-transmitting plate 3, and then the electric telescopic rod 19 drives the mounting plate 5 to descend, so that the collection tube 6 contacts the potatoes. As the collection tube 6 continues to descend, the cutting ring 8 cuts the potatoes, and only potatoes with the same area as the collection tube are retained. At this time, the light source lamp 4 at the bottom of the collection box 2 starts to work to supplement the light source, and the light source is transmitted to the potatoes through the light-transmitting plate 3. At the same time, the hyperspectral imager 7 collects images of the potatoes. After the spectral images of the collected potatoes are processed by the analyzer and the processor, the corresponding images and data intervals are obtained on the collection computer 23. In the subsequent test, if the values ​​of the spectral images of the potato slices are all within the value interval, it means that the potatoes are healthy;

[0043] S2: Model establishment: After the image is collected, the anthocyanin, starch and protein contents of the potatoes used for collecting the images are measured, and the data range of the anthocyanin, starch and protein contents of normal potatoes is first determined, and then the data range of the anthocyanin, starch and protein contents of potatoes infected with ring rot is determined. The data of healthy potatoes and infected potatoes are compared to determine the impact of potato infection with ring rot on the anthocyanin, starch and protein contents. The spectral image of healthy potatoes and the anthocyanin, starch and protein data of healthy potatoes are classified, and the spectral image of infected potatoes and the anthocyanin, starch and protein data of infected potatoes are classified. Based on the anthocyanin, starch and protein data of the spectral image, a regression model of potatoes infected with ring rot and not infected with ring rot and a classification model of the infection degree of ring rot are established. By comparing the spectral image with the anthocyanin, starch and protein data, it is possible to determine whether the potato is infected, and the obtained model is written into the analysis module of the acquisition computer 23;

[0044] S3: Model verification: After the model is established, potato slices are taken for image acquisition. The acquired images are automatically input into the regression model and the classification model for analysis and processing. The acquisition computer 23 displays the analysis results to determine whether the potato suffers from ring rot. Then, the potato used for verification is subjected to detection processing such as staining. The detection structure is compared with the model verification result to observe whether the two results are consistent, and the accuracy of the model is reversely verified. If the results are consistent, it means that the model is successfully established. If the results are inconsistent, adjustments are continued;

[0045] S4: Application of the model: After the model is verified to be able to operate and work normally, the potatoes can be directly inspected. When inspecting potatoes, open the box cover 20 and place the potatoes on the light-transmitting plate 3. Then the electric telescopic rod 19 drives the mounting plate 5 to descend so that the collecting tube 6 contacts the potatoes. As the collecting tube 6 continues to descend, the cutting ring 8 cuts the potatoes and only retains potatoes with the same area as the collecting tube. At this time, the light source lamp 4 at the bottom of the collecting box 2 starts to work to supplement the light source. The light source is transmitted to the potatoes through the light-transmitting plate 3. At the same time, the hyperspectral imager 7 collects images of the potatoes. After the collected images are processed by the analysis module, the corresponding images and detection results are obtained on the collection computer 23.

[0046] Working principle of the present invention:

[0047] Before the test, a number of healthy potatoes and diseased potatoes are selected for slicing, and the sliced ​​potatoes are classified for image collection. The potatoes to be tested are sliced, and the slicing parts include the inside and the outer surface of the potato. After slicing, the box cover 20 is opened, and the potato is placed on the light-transmitting plate 3. Then the electric telescopic rod 19 drives the mounting plate 5 to descend, so that the collection tube 6 contacts the potato. As the collection tube 6 continues to descend, the cutting ring 8 cuts the potato, and only the potato with the same area as the collection tube is retained. At this time, the light source lamp 4 at the bottom of the collection box 2 starts to work to supplement the light source, and the light source is transmitted to the potato through the light-transmitting plate 3. At the same time, the hyperspectral imager 7 collects images of the potato. After the collected image is processed by the analyzer and the processor, the corresponding image is obtained on the collection computer 23;

[0048] Model establishment: after image acquisition, the anthocyanin, starch and protein contents of the potatoes used for image acquisition are measured, firstly the data range of the anthocyanin, starch and protein contents of normal potatoes is determined, then the data range of the anthocyanin, starch and protein contents of potatoes infected with ring rot is determined, the data of healthy potatoes and infected potatoes are compared, the effect of potato infection with ring rot on the anthocyanin, starch and protein contents is determined, and based on this data, a regression model of potatoes infected with ring rot and not infected with ring rot and a classification model of the infection degree of ring rot are established, and the obtained model is written into the analysis module of the acquisition computer 23;

[0049] After the model is established, potato slices are taken for image acquisition. The acquired images are automatically input into the regression model and the classification model for analysis and processing. The acquisition computer 23 displays the analysis results to determine whether the potato suffers from ring rot. Then, the potato used for verification is dyed and other detection processes are performed. The detection structure is compared with the model verification result to observe whether the two results are consistent, and the accuracy of the model is reversely verified. If the results are consistent, it means that the model is successfully established. If the results are inconsistent, adjustments are continued;

[0050] After the model is verified to be able to operate and work normally, the potatoes can be directly tested. When testing potatoes, the box cover 20 is opened, and the potatoes are placed on the light-transmitting plate 3. Then the electric telescopic rod 19 drives the mounting plate 5 to descend, so that the collection tube 6 contacts the potatoes. As the collection tube 6 continues to descend, the cutting ring 8 cuts the potatoes, and only potatoes with the same area as the collection tube are retained. At this time, the light source lamp 4 at the bottom of the collection box 2 starts to work to supplement the light source. The light source is transmitted on the potatoes through the light-transmitting plate 3. At the same time, the hyperspectral imager 7 collects images of the potatoes. After the collected images are processed by the analysis module, the corresponding images and detection results are obtained on the collection computer 23.

[0051] During operation, the potato to be inspected is sliced, and then the box cover 20 is opened, and the potato to be inspected is laid flat on the light-transmitting plate 3 according to the corresponding position of the collection tube 6, and the light source lamp 4 is turned on at the same time. At this time, the mounting plate 5 moves downward under the action of the control component and is inserted into the collection box 2. As the cutting ring 8 contacts the potato, the irregular part of the potato edge is cut off, and the remaining part is completely merged with the collection tube 6. Compared with the traditional image acquisition device, the potato detection area is limited by the collection tube 6, which avoids the light transmission problem caused by the irregular edge of the potato, and the spectral image acquisition is more accurate and without interference. At the same time, the collection box 2 blocks the external light to prevent the light source in the external wall environment from affecting the collection of the spectral image, and the spectral image acquisition is more accurate.

[0052] When collecting spectral images of potatoes, the servo motor 12 drives the threaded rod 11 to rotate, which can drive the mounting block 9 to move up and down, thereby adjusting the distance between the light-transmitting plate 3 and the light source lamp 4, thereby adjusting the light source intensity according to the thickness of the potato slices, and the operation is more convenient.

[0053] The above description is only a preferred specific implementation manner of the present invention, but the protection scope of the present invention is not limited thereto. Any technician familiar with the technical field can make equivalent replacements or changes according to the technical scheme and inventive concept of the present invention within the technical scope disclosed by the present invention, which should be covered by the protection scope of the present invention.

Claims

1. A potato diffuse transmission hyperspectral image acquisition device, comprising a bottom plate (1), characterized in that: Support legs are welded around the bottom outer wall of the bottom plate (1); a collection box (2) is installed on one end of the top outer wall of the bottom plate (1); the collection box (2) is made of opaque material; the collection box (2) is compatible with the bottom plate (1); a control component is installed on one end of the bottom plate (1) close to the collection box (2); the control component is connected to a collection component; the collection component is compatible with the control component; a light-transmitting plate (3) is slidably installed on the inner wall of the collection box (2); the light-transmitting plate (3) is compatible with the collection box (2); The inner wall at the bottom of the collection box (2) is provided with light source lamps (4) distributed at equal distances, the light-transmitting plate (3) is adapted to the collection component, the collection component comprises a mounting plate (5), the bottom of the mounting plate (5) is provided with collection tubes (6) distributed at equal distances, the top inner wall of the collection tube (6) is provided with a hyperspectral imager (7), the hyperspectral imager (7) is adapted to the collection tube (6), the outer wall at the bottom of the collection tube (6) is provided with a cutting ring (8), the diameter of the cutting ring (8) being the same as the diameter of the collection tube (6).

2. The potato diffuse transmission hyperspectral image acquisition device according to claim 1, characterized in that: Mounting blocks (9) are slidably mounted on the inner walls of both sides of the collection box (2); the light-transmitting plate (3) is fixedly connected to the mounting blocks (9); the light-transmitting plate (3) and the mounting blocks (9) are adapted to each other; adjustment grooves (10) are provided on the inner walls of both sides of the collection box (2); the mounting blocks (9) are slidably connected to the inner walls of the adjustment grooves (10); a threaded rod (11) is rotatably mounted on the inner walls of the adjustment grooves (10); the length of the threaded rod (11) is the same as that of the adjustment grooves (10); and the mounting blocks (9) and the threaded rod (11) are rotatably connected via threads.

3. A potato diffuse transmission hyperspectral image acquisition device according to claim 2, characterized in that: The bottom of the inner walls on both sides of the collection box (2) are provided with driving grooves, and the inner walls of the driving grooves are provided with servo motors (12), and the output shaft of the servo motor (12) is connected to the threaded rod (11).

4. The potato diffuse transmission hyperspectral image acquisition device according to claim 3, characterized in that: The outer walls on both sides of the light-transmitting plate (3) are provided with fixing blocks (13), the outer walls on both sides of the adjacent mounting blocks (9) are provided with fixing grooves (14) distributed at equal distances, the fixing blocks (13) are adapted to the fixing grooves (14), the outer walls of the fixing blocks (13) and the inner walls at the bottoms of the fixing grooves (14) are provided with fixing holes (15), the inner walls of the fixing holes (15) are threadedly mounted with fixing bolts (16), and the length of the fixing bolts (16) is greater than the thickness of the fixing blocks (13).

5. The potato diffuse transmission hyperspectral image acquisition device according to claim 4, characterized in that: The control assembly comprises a mounting rod (17) welded to one end of the top outer wall of the base plate (1), a mounting frame (18) is welded to the top of the mounting rod (17), the mounting frame (18) is a "T"-shaped structure, a mounting hole is formed on the top outer wall of the mounting frame (18), an electric telescopic rod (19) is mounted on the inner wall of the mounting hole, and the end of the electric telescopic rod (19) is connected to the mounting plate (5).

6. A potato diffuse transmission hyperspectral image acquisition device according to claim 5, characterized in that: A box cover (20) is slidably mounted on the top of the collection box (2), a collection slot (21) is provided on the top of the box cover (20), the collection slot (21) is matched with the mounting plate (5), a control block (22) is provided at one end of the top outer wall of the box cover (20), sliding slots are provided at both ends of the top outer wall of the collection box (2), and sliding blocks are welded at both ends of the bottom outer wall of the box cover (20), and the sliding blocks are slidably connected to the inner wall of the sliding slot.

7. The potato diffuse transmission hyperspectral image acquisition device according to claim 6, characterized in that: A collection computer (23) is installed at one end of the top outer wall of the base plate (1); the collection computer (23) is connected to a data processor via a wireless network; the data processor is connected to an image analyzer via a wireless network; and the image analyzer is connected to a hyperspectral imager (7) via a wireless network.

8. The potato diffuse transmission hyperspectral image acquisition device according to claim 7, characterized in that: A control panel (24) is installed on the top outer wall of the base plate (1) near the acquisition computer (23), and a control button (25) is arranged on the outer wall of the control panel (24). The control button (25) is connected to the electrical components via a wire.

9. The potato diffuse transmission hyperspectral image acquisition device according to claim 8, characterized in that: The top outer wall of the bottom plate (1) is provided with a mounting groove (26), the mounting groove (26) being compatible with the collection box (2), and the depth of the mounting groove (26) being less than the height of the collection box (2).

10. A potato diffuse transmission hyperspectral image acquisition method, using a potato diffuse transmission hyperspectral image acquisition device as claimed in claim 9, characterized in that: The method comprises the following steps: S1: Image acquisition: Before testing, a number of healthy potatoes and diseased potatoes are selected for slicing, and the sliced ​​potatoes are classified for image acquisition. The potatoes to be tested are sliced. After slicing, the box cover (20) is opened, and the potatoes are placed on the light-transmitting plate (3). Then, the electric telescopic rod (19) drives the mounting plate (5) to descend, so that the collection tube (6) contacts the potatoes. As the collection tube (6) continues to descend, the cutting ring (8) cuts the potatoes, and only potatoes with the same area as the collection tube are retained. At this time, the light source lamp (4) at the bottom of the collection box (2) starts to work to supplement the light source. The light source is transmitted through the light-transmitting plate (3) onto the potatoes. At the same time, the hyperspectral imager (7) collects images of the potatoes. The collected spectral images of the potatoes are processed by the analyzer and the processor, and the corresponding images and data intervals are obtained on the collection computer (23). In subsequent testing, if the values ​​of the spectral images of the potato slices are all within the value interval, it means that the potatoes are healthy; S2: Model establishment: After image acquisition, the anthocyanin, starch and protein content of the potato used for image acquisition is measured. The data range of the anthocyanin, starch and protein content of normal potato is first determined, and then the data range of the anthocyanin, starch and protein content of potato infected with ring rot is determined. The data of healthy potato and infected potato are compared to determine the effect of potato infection with ring rot on the anthocyanin, starch and protein content. The spectral image of healthy potato and the anthocyanin, starch and protein data of healthy potato are classified. The spectral image of infected potato and the anthocyanin, starch and protein data of infected potato are classified. Based on the anthocyanin, starch and protein data of spectral image, a regression model of potato infected with ring rot and not infected with ring rot and a classification model of the infection degree of ring rot are established. By comparing the spectral image with the anthocyanin, starch and protein data, it is possible to judge whether the potato is infected, and the obtained model is written into the analysis module of the acquisition computer (23); S3: Model verification: After the model is established, potato slices are taken for image acquisition. The acquired images are automatically input into the regression model and the classification model for analysis and processing. The acquisition computer (23) displays the analysis results to determine whether the potato suffers from ring rot. Then, the potatoes used for verification are stained and tested. The test structure is compared with the model verification result to observe whether the two results are consistent. The accuracy of the model is reversely verified. If the results are consistent, it means that the model is successfully established. If the results are inconsistent, adjustments are continued. S4: Application of the model: After the model is verified to be able to operate and work normally, the potato can be directly inspected. When inspecting the potato, the box cover (20) is opened, and the potato is placed on the light-transmitting plate (3). Then the electric telescopic rod (19) drives the mounting plate (5) to descend, so that the collection tube (6) contacts the potato. As the collection tube (6) continues to descend, the cutting ring (8) cuts the potato, and only the potato with the same area as the collection tube is retained. At this time, the light source lamp (4) at the bottom of the collection box (2) starts to work to supplement the light source. The light source is transmitted to the potato through the light-transmitting plate (3). At the same time, the hyperspectral imager (7) collects images of the potato. After the collected image is processed by the analysis module, the corresponding image and detection result are obtained on the collection computer (23).

Images