Leguminous crop root nodule feature recognition device and recognition method
By designing a composite light source system and image processing system in legume crops, combining rulers and neural networks, rapid and accurate nodule tumor feature recognition is achieved, solving the problem of low efficiency in the existing technology, and improving nodule count and size measurement efficiency.
Patent Information
- Application Number
- CN202411093204.4
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-08-09
- Publication Date
- 2024-11-19
AI Technical Summary
The prior art is difficult to quickly and accurately identify the nodule tumor characteristic information of legumes, resulting in low efficiency in nodule counting and size measurement, especially in legumes, which are large and difficult to distinguish.
A nodule feature recognition device including a composite light source system, an image capture system and an image processing system is designed. Using the combination of blue light and white light surface light sources, the characteristic spectral image of the nodule is extracted through an image processing algorithm, and the standard and neural network are combined for accurate identification.
Fast and accurate nodule tumor feature recognition is achieved, improving the efficiency of nodule tumor counting and size measurement in legume crops, with high accuracy, portability and losslessness.
Smart Images

Figure CN118980648A_ABST
Abstract
Description
Technical Field
[0001] The invention relates to the fields of plant science and agricultural biotechnology, and in particular to a device and method for identifying root nodule characteristics of leguminous crops. Background Art
[0002] Nodules are symbiotic nitrogen-fixing structures formed between the roots of legumes and rhizobia. Nodules are formed when rhizobia living in the soil invade the roots. The relationship between rhizobia and legumes is a mutually beneficial symbiotic relationship. Improving soil structure and increasing soil fertility through biological nitrogen fixation is an ideal way to restore the ecosystem. The number, size and color of nodules are important indicators of legume health, and dynamic monitoring of nodules is very important for evaluating plant status.
[0003] The existing method for counting and measuring nodules in the art is to dig out the roots, identify the nodules with the naked eye, and count and measure them manually, which makes the nodule counting and activity determination boring and time-consuming, and is greatly affected by personnel and external factors. Especially for leguminous crops such as peanuts and soybeans, due to the large number of nodules and the difficulty in distinguishing them from the roots of crops, the existing technology lacks effective means for identifying nodule characteristics. Therefore, how to quickly and accurately identify the nodule characteristic information of leguminous crops is of great significance for evaluating the apparent information such as the number and size of nodules of leguminous crops, as well as for promoting the development of biological nitrogen fixation technology. Summary of the invention
[0004] The technical problem to be solved by the present invention is to provide a device and method for identifying the characteristics of leguminous crop nodules, which can quickly and accurately identify the characteristic information of the nodules of leguminous crops and greatly improve the efficiency of counting and measuring the size of leguminous crop nodules.
[0005] In order to solve the above technical problems, the technical solution adopted by the present invention is:
[0006] 1. A device for identifying the characteristics of nodules of leguminous crops
[0007] The present invention provides a device for identifying root nodules of leguminous crops, comprising: a device housing 1, a composite light source system, an image shooting system and an image processing system 8;
[0008] The composite light source system comprises a blue light surface light source 2 arranged on the bottom surface of the inner cavity of the device housing 1 and a white light surface light source 6 arranged on the top surface of the inner cavity of the device housing 1;
[0009] The image capture system includes a ruler 3 with a standard color card and an imaging device 7, wherein the ruler 3 is arranged on the upper surface of the blue light source 2 on the bottom surface of the inner cavity of the device housing 1, and the imaging device 7 is installed in a shooting hole opened in the center of the top of the device housing 1, and a filter 5 and a polarizer 4 are arranged in sequence at the bottom of the lens of the imaging device 7;
[0010] The image processing system 8 is electrically connected to the imaging device 7 , and an image processing algorithm runs in the image processing system 8 .
[0011] Preferably, the light emission band of the blue light surface light source 2 is adapted to the absorption spectrum peak of leghemoglobin in the root nodules of the sample to be identified, and the illumination value of the blue light surface light source 2 is 3 to 5 times that of the white light surface light source 6 .
[0012] Preferably, the blue light surface light source 2 and the white light surface light source 6 are both provided with a light source adjustment device for controlling the opening, closing and illumination value adjustment of the corresponding light sources.
[0013] Preferably, the wavelength of the polarizer 4 is 0.9L min ~1.1L max , used to filter the reflected light, scattered light and flash light from water droplets on the root surface of the sample to be identified.
[0014] Among them, the L min and L max They are respectively the minimum wavelength value and the maximum wavelength value of the light-emitting band of the blue light surface light source 2.
[0015] Preferably, the filtering wavelength band of the filter 5 is adapted to the light emitting wavelength band of the blue light surface light source 2, and is used to filter the background light and enhance the saturation of the blue light surface light source.
[0016] Preferably, the sample to be identified is placed on the upper surface of the blue light source 2 on the bottom surface of the inner cavity of the device housing 1 and is located on one side of the scale 3;
[0017] The imaging device 7 is used to obtain the characteristic spectrum image of the root system of the sample to be identified and the scale 3, and send the image data to the image processing system 8, and the pixel ratio of the root system of the sample to be identified and the scale 3 in the characteristic spectrum image remains consistent.
[0018] 2. A method for identifying root nodules of leguminous crops
[0019] Based on the same inventive concept, the present invention also provides a method for identifying leguminous crop nodule characteristics, based on the leguminous crop nodule characteristic identification device as described above, specifically comprising the following steps:
[0020] S1, preparation of samples to be identified: dig out the roots of the samples to be identified completely from the soil, and remove the soil and impurities on the roots in clean water;
[0021] S2, root system pretreatment: cut the lateral roots that are not in the same plane as the main root, lay the cut main root and lateral roots completely flat on the upper surface of the blue light source, and place a ruler with a standard color card on one side of the sample root system;
[0022] S3, composite light source setting: turning on the blue light source on the bottom surface of the inner cavity of the device shell and the white light source on the top surface, and adjusting the illumination of the blue light source and the white light source to a preset value;
[0023] S4, taking root system images: acquiring characteristic spectral images of the root system and the scale of the sample to be identified through an imaging device, and sending the image data to the image processing system;
[0024] S5, image processing and identification: using an image processing algorithm to extract the phenotypic characteristics of the root nodules of the sample to be identified in the characteristic spectral image, and based on the phenotypic characteristics of the nodules combined with the identification model, obtaining the size and quantity parameters of the nodules.
[0025] Furthermore, the image processing algorithm specifically includes:
[0026] S11, performing cropping and color size normalization preprocessing on the characteristic spectrum image;
[0027] S12, performing binarization processing on the pre-processed characteristic spectrum image to highlight the contour features of the nodule so as to perform nodule image segmentation and extraction;
[0028] S13, using a neural network to build a recognition model to accurately identify target nodules in the segmented and extracted nodule images;
[0029] S14, identifying the size and number parameters of nodules based on the single pixel size of the ruler combined with the minimum circumscribed circle estimation method.
[0030] The characteristic spectrum image is cropped to retain only the part of the root system image of the sample to be identified that contains the nodule and the part where the scale is located;
[0031] The color size normalization preprocessing is specifically to obtain the minimum brightness value D of the characteristic spectrum image min and the maximum brightness D max , and adjust the brightness value D of each pixel in the characteristic spectrum image to:
[0032] D=(D0-D min ) / (D max -D min )
[0033] Where D0 is the brightness value of each pixel before adjustment.
[0034] Furthermore, the size and quantity parameters of the identified nodules specifically include:
[0035] 1) According to the accurately identified target nodule image, the overall pixel number of the target nodule image is obtained;
[0036] 2) calculating the overall actual size of the target nodule image based on the actual size corresponding to a single pixel of the ruler and the overall number of pixels of the target nodule image;
[0037] 3) The average size of a single nodule is estimated by using the minimum circumscribed circle of image recognition, and the number parameter of the nodule is obtained by dividing the overall actual size by the average size of the single nodule.
[0038] Compared with the prior art, the present invention has the following main advantages:
[0039] 1. The present invention has a compact overall structure and convenient installation and arrangement through the reasonable arrangement of the device housing, the composite light source system, and the image shooting system. It can directly measure samples in the field, and cooperate with the image processing system to realize the rapid acquisition and accurate feature recognition of the phenotypic image spectrum of leguminous crop nodules, which can greatly improve the efficiency of leguminous crop nodule counting and size measurement.
[0040] 2. The present invention can expand the color contrast between the root system and the nodules, enhance the nodule hierarchy, and facilitate the separation of nodules and root system characteristics through the top white light surface light source and the bottom blue light surface light source. By setting filters and polarizers, the spectral characteristics can be improved and the interference of the environment on the image can be reduced.
[0041] 3. The present invention utilizes the spectral characteristics of leghemoglobin in the nodules of leguminous crops, adopts a composite light source system to obtain characteristic spectral images of nodules, uses characteristic spectral information to locate the position of nodules, and combines image processing algorithms to dually locate nodules. It can quickly distinguish the roots, soil particles and nodule image characteristics of leguminous crops, clearly and accurately identify the phenotypic characteristics (size, quantity) of nodules, and thus realize efficient nodule counting, size measurement and evaluation, with the advantages of high accuracy, rapidity, portability and non-destructiveness. BRIEF DESCRIPTION OF THE DRAWINGS
[0042] Figure 1 Schematic diagram of the structure of a device for identifying nodules of leguminous crops according to an embodiment of the present invention;
[0043] Figure 2 Flow chart of a method for identifying nodule characteristics of leguminous crops in an embodiment of the present invention;
[0044] Figure 3 is a flowchart of an image processing algorithm in an embodiment of the present invention;
[0045] Figure 4 Schematic diagram of nodules at the roots of peanut crops in an embodiment of the present invention.
[0046] In the figure: 1. Device housing; 2. Blue light surface light source; 3. Ruler with standard color card; 4. Polarizer; 5. Filter; 6. White light surface light source; 7. Imaging device; 8. Image processing system. DETAILED DESCRIPTION
[0047] In order to make the purpose, technical solutions and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention. In addition, the technical features involved in the various embodiments of the present invention described below can be combined with each other as long as they do not conflict with each other.
[0048] It should be pointed out that, according to the needs of implementation, the various steps / components described in this application can be split into more steps / components, and two or more steps / components or partial operations of steps / components can be combined into new steps / components to achieve the purpose of the present invention.
[0049] Embodiment 1: This embodiment provides a device for identifying root nodules of leguminous crops. Figure 1 As shown, it includes: a device housing 1, a composite light source system, an image capturing system and an image processing system 8;
[0050] The composite light source system comprises a blue light surface light source 2 arranged on the bottom surface of the inner cavity of the device housing 1 and a white light surface light source 6 arranged on the top surface of the inner cavity of the device housing 1;
[0051] The image capture system includes a ruler 3 with a standard color card and an imaging device 7, wherein the ruler 3 is arranged on the upper surface of the blue light source 2 on the bottom surface of the inner cavity of the device housing 1, and the imaging device 7 is installed in a shooting hole opened in the center of the top of the device housing 1, and a filter 5 and a polarizer 4 are arranged in sequence at the bottom of the lens of the imaging device 7;
[0052] The image processing system 8 is electrically connected to the imaging device 7 , and an image processing algorithm runs in the image processing system 8 .
[0053] Furthermore, the light emission band of the blue light surface light source 2 is adapted to the absorption spectrum peak of leghemoglobin in the root nodules of the sample to be identified, and the illumination value of the blue light surface light source 2 is 3 to 5 times that of the white light surface light source 6 .
[0054] Furthermore, the blue light surface light source 2 and the white light surface light source 6 are both provided with a light source adjustment device for controlling the opening, closing and illumination value adjustment of the corresponding light sources.
[0055] Furthermore, the wavelength of the polarizer 4 is 0.9L min ~1.1L max, used to filter the reflected light, scattered light and flash light from water droplets on the root surface of the sample to be identified.
[0056] Among them, the L min and L max They are respectively the minimum wavelength value and the maximum wavelength value of the light-emitting band of the blue light surface light source 2.
[0057] Furthermore, the filter band of the filter 5 is adapted to the light emitting band of the blue light surface light source 2, and is used to filter the background light and enhance the saturation of the blue light surface light source.
[0058] Furthermore, the sample to be identified is placed on the upper surface of the blue light source 2 on the bottom surface of the inner cavity of the device housing 1 and is located on one side of the scale 3;
[0059] The imaging device 7 is used to obtain the characteristic spectrum image of the root system of the sample to be identified and the scale 3, and send the image data to the image processing system 8, and the pixel ratio of the root system of the sample to be identified and the scale 3 in the characteristic spectrum image remains consistent.
[0060] Embodiment 2: This embodiment provides a device for identifying the characteristics of leguminous crop nodules, which is mainly composed of a black device housing 1, a composite light source system, an image shooting system and an image processing system 8, wherein an image processing algorithm is running in the image processing system.
[0061] The composite light source system includes a white light surface light source 6 arranged at the top of the inner cavity of the device housing 1 and a blue light surface light source 2 arranged at the bottom; the image capture system includes an imaging device 7 installed in the small shooting hole at the top of the device housing 1 and a ruler 3 with a standard color card placed above the blue light surface light source 2 at the bottom, and the bottom of the lens of the imaging device 7 is equipped with a filter 5 and a polarizer 4 in sequence;
[0062] Furthermore, the wavelength of the blue light surface light source is 400-450nm, the illumination of the blue light surface light source is 1200-2000lux, and the illumination of the white light surface light source is 400-800lux. The illumination value of the blue light surface light source is 3-5 times that of the white light surface light source. By using upper and lower fill light, the layering and color level of the nodules are enhanced, the contrast between the root system and the nodules is expanded, and the separation of the nodule and root characteristics is achieved.
[0063] Furthermore, the filter is a high-transmittance filter of 420-450nm, and the use of the filter can improve the blue light nodule root characteristics and reduce background interference;
[0064] Furthermore, the wavelength of the polarizer is 400nm to 500nm; the polarizer can effectively filter out the interference of reflected light, scattered light and flash light from water droplets on the surface of the root system on the image.
[0065] Furthermore, the ruler with the standard color card is 35-55 cm long and 2-3.5 cm wide, and is used to normalize the size of root nodules in the image.
[0066] Furthermore, the device uses blue light in a specific band to obtain characteristic spectral images of nodules, combines image processing algorithms to extract nodule phenotypic characteristic information, counts indicators such as the number, color, and size of nodules, and uses differences in the spectral characteristics of roots and nodules to achieve nodule counting evaluation.
[0067] Embodiment 3, based on the same inventive concept, this embodiment also provides a method for identifying leguminous crop nodule characteristics, based on the leguminous crop nodule characteristic identification device as described above, such as Figure 2 As shown, the specific steps include:
[0068] S1, preparation of samples to be identified: dig out the roots of the samples to be identified from the soil completely, and wash them in clean water to remove excess soil and impurities;
[0069] S2, root system pretreatment: use scissors to cut the lateral roots that are not in the same plane from the junction of the main root, lay all the cut roots and main roots flat on the blue light surface source board, and place the ruler with the standard color card on the blue light surface source board;
[0070] S3, composite light source setting: turn on the bottom blue light surface light source and the top white light surface light source, and adjust the current so that the illumination of the blue light surface light source is 3-5 times that of the white light source;
[0071] S4, taking root system images: installing the imaging device at the top shooting port to obtain the root system spectral image;
[0072] S5, image processing and identification: obtain characteristic spectral images of root nodules in a specific blue light band, use image processing algorithms to extract nodule phenotypic characteristics (size, number), and use the differences in spectral characteristics of roots and nodules to count nodules.
[0073] Furthermore, the image processing algorithm, such as Figure 3 As shown, specifically including:
[0074] S11, performing cropping and color size normalization preprocessing on the root nodule spectral image;
[0075] S12, performing binarization processing on the preprocessed image by using a global threshold method or an adaptive threshold method, so as to highlight the contour and features of the nodule and realize the nodule segmentation;
[0076] S13, using neural networks such as YOLO V8 to build a recognition model and accurately identify target nodules;
[0077] S14, calculating the minimum circumscribed circle to estimate the nodule size, combining the recognition model to achieve root nodule segmentation and detection, and obtaining characteristic parameters such as nodule size and number.
[0078] Furthermore, the ruler with the standard color card is also used to identify the actual color of the nodules in the root system of the sample to be identified, and the specific process includes:
[0079] S21, turn off the blue light source on the bottom surface of the inner cavity of the device shell, and turn on the white light source on the top surface
[0080] S22, acquiring a color image of the sample nodule to be identified and the ruler with the standard color card through an imaging device, and sending the color image data to the image processing system;
[0081] S23, comparing the actual color of the standard color card with the color of the standard color card in the color image, and determining a color deviation of the color image acquired by the imaging device;
[0082] S24, correcting the color of the nodules in the color image according to the color deviation to obtain the actual color of the nodules of the sample to be identified.
[0083] Based on obtaining the size and number of nodules in the root system of the sample to be identified through image processing algorithms, this solution further calibrates the color deviation of the imaging device through a ruler with a standard color card to obtain the actual color of the nodules, thereby achieving an overall evaluation of the nodule count, size, and color.
[0084] Embodiment 4, as Figure 4 As shown, this embodiment takes the root system of peanut crops as an example, and uses the above-mentioned nodule feature recognition method to perform nodule counting and size measurement:
[0085] Step 1: Sample preparation: dig out a peanut plant with complete root system from the soil, wash the peanut roots in clean water, and remove excess soil and impurities attached to the root surface;
[0086] Step 2: Root pretreatment: Find the lateral roots that are not in the same plane in the peanut root system, cut them from the junction of the main root with scissors, lay all the cut roots and main roots flat on the blue light background plate of the image acquisition device, and put in a ruler with a standard color card;
[0087] Step 3: Composite light source settings: black shell size: 45cm*45cm, blue light surface light source size: 40*29cm, power 25w, filter: wavelength 420~450nm, polarizer wavelength 400nm~500nm, white light surface light source size: 40*29cm, power 15w, scale size 35cm*3cm, the distance between the blue light surface light source and the white light surface light source is 60cm, turn on the bottom blue light surface light source and the top white light surface light source, adjust the current so that the illumination of the blue light surface light source is 1600lux, and the illumination of the white light source is 800lux;
[0088] Step 4: Take root system images: Install the imaging device at the top shooting port to take spectral images of the entire root system and taproot of peanut;
[0089] Step 5: Image processing and identification: Using computer image processing algorithms, nodule features are extracted from the root nodule images. By combining nodule feature information data with an identification model, it is concluded that the number of nodules in the sample is 530, and the radius range is 0.1429 mm to 0.3858 mm.
[0090] Furthermore, all parts involved in this application that are not described in detail are the same as the prior art or are implemented using the prior art.
[0091] In summary:
[0092] 1. The present invention has a compact overall structure and convenient installation and arrangement through the reasonable arrangement of the device housing, the composite light source system, and the image shooting system. It can directly measure samples in the field, and cooperate with the image processing system to realize the rapid acquisition and accurate feature recognition of the phenotypic image spectrum of leguminous crop nodules, which can greatly improve the efficiency of leguminous crop nodule counting and size measurement.
[0093] 2. The present invention can expand the color contrast between the root system and the nodules, enhance the nodule hierarchy, and facilitate the separation of nodules and root system characteristics through the top white light surface light source and the bottom blue light surface light source. By setting filters and polarizers, the spectral characteristics can be improved and the interference of the environment on the image can be reduced.
[0094] 3. The present invention utilizes the spectral characteristics of leghemoglobin in the nodules of leguminous crops, adopts a composite light source system to obtain characteristic spectral images of nodules, uses characteristic spectral information to locate the position of nodules, and combines image processing algorithms to dually locate nodules. It can quickly distinguish the roots, soil particles and nodule image characteristics of leguminous crops, clearly and accurately identify the phenotypic characteristics (size, quantity) of nodules, and thus realize efficient nodule counting, size measurement and evaluation, with the advantages of high accuracy, rapidity, portability and non-destructiveness.
[0095] It will be easily understood by those skilled in the art that the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions and improvements made within the spirit and principles of the present invention should be included in the protection scope of the present invention.
Claims
1. A device for identifying root nodules of leguminous crops, characterized in that: include: Device housing (1), composite light source system, image capturing system and image processing system (8); The composite light source system comprises a blue light surface light source (2) arranged on the bottom surface of the inner cavity of the device housing (1) and a white light surface light source (6) arranged on the top surface of the inner cavity of the device housing (1); The image capture system comprises a ruler (3) with a standard color card and an imaging device (7), wherein the ruler (3) is arranged on the upper surface of a blue light surface light source (2) on the bottom surface of an inner cavity of a device housing (1), and the imaging device (7) is installed in a shooting hole opened in the center of the top of the device housing (1), and a filter (5) and a polarizer (4) are arranged in sequence at the bottom of the lens of the imaging device (7); The image processing system (8) is electrically connected to the imaging device (7), and an image processing algorithm runs in the image processing system (8).
2. The device for identifying root nodules of leguminous crops according to claim 1, characterized in that: The light emission band of the blue light surface light source (2) is compatible with the absorption spectrum peak of leghemoglobin in the root nodules of the sample to be identified, and the illumination value of the blue light surface light source (2) is 3 to 5 times the illumination value of the white light surface light source (6).
3. The device for identifying root nodules of leguminous crops according to claim 2, characterized in that: The blue light surface light source (2) and the white light surface light source (6) are both provided with a light source adjustment device for controlling the opening and closing of the corresponding light sources and adjusting the illumination value.
4. The device for identifying root nodules of leguminous crops according to claim 1, characterized in that: The wavelength of the polarizer (4) is 0.9L min ~1.1L max , used to filter the reflected light, scattered light and flash light from water droplets on the root surface of the sample to be identified. Among them, the L min and L max They are respectively the minimum wavelength value and the maximum wavelength value of the light-emitting band of the blue light surface light source (2).
5. The device for identifying root nodules of leguminous crops according to claim 4, characterized in that: The filter (5) has a filtering wavelength band that matches the light emitting wavelength band of the blue light surface light source (2), and is used to filter background light and enhance the saturation of the blue light surface light source.
6. The device for identifying root nodules of leguminous crops according to claim 2, characterized in that: The sample to be identified is placed on the upper surface of the blue light source (2) on the bottom surface of the inner cavity of the device housing (1) and is located on one side of the scale (3); The imaging device (7) is used to obtain characteristic spectrum images of the root system of the sample to be identified and the scale (3), and send the image data to the image processing system (8), and the pixel ratio of the root system of the sample to be identified and the scale (3) in the characteristic spectrum image remains consistent.
7. A method for identifying leguminous crop nodule characteristics, based on the leguminous crop nodule characteristics identification device according to any one of claims 1 to 6, characterized in that: The steps include: S1, preparation of samples to be identified: dig the roots of the samples to be identified completely from the soil, and remove the soil and impurities on the roots in clean water; S2, root system pretreatment: cut the lateral roots that are not in the same plane as the main root, lay the cut main root and lateral roots completely flat on the upper surface of the blue light source, and place a ruler with a standard color card on one side of the sample root system; S3, composite light source setting: turning on the blue light source on the bottom surface of the inner cavity of the device shell and the white light source on the top surface, and adjusting the illumination of the blue light source and the white light source to a preset value; S4, taking root system images: acquiring characteristic spectrum images of the root system and the scale of the sample to be identified through an imaging device, and sending the image data to the image processing system; S5, image processing and identification: using an image processing algorithm to extract the phenotypic characteristics of the root nodules of the sample to be identified in the characteristic spectral image, and based on the phenotypic characteristics of the nodules combined with the identification model, obtaining the size and quantity parameters of the nodules.
8. The method for identifying nodule characteristics of leguminous crops according to claim 7, characterized in that: The image processing algorithm specifically includes: S11, performing cropping and color size normalization preprocessing on the characteristic spectrum image; S12, performing binarization processing on the pre-processed characteristic spectrum image to highlight the contour features of the nodule so as to perform nodule image segmentation and extraction; S13, using a neural network to build a recognition model to accurately identify target nodules in the segmented and extracted nodule images; S14, identifying the size and number parameters of nodules based on the single pixel size of the ruler combined with the minimum circumscribed circle estimation method.
9. The method for identifying nodule characteristics of leguminous crops according to claim 8, characterized in that: The characteristic spectrum image is cropped to retain only the part of the root system image of the sample to be identified that contains the nodule and the part where the scale is located; The color size normalization preprocessing is specifically to obtain the minimum brightness value D of the characteristic spectrum image min and the maximum brightness D max , and adjust the brightness value D of each pixel in the characteristic spectrum image to: D=(D0-D min ) / (D max -D min ) Where D0 is the brightness value of each pixel before adjustment.
10. The method for identifying nodule characteristics of leguminous crops according to claim 9, characterized in that: The size and quantity parameters of the identified nodules specifically include: 1) According to the accurately identified target nodule image, the overall pixel number of the target nodule image is obtained; 2) calculating the overall actual size of the target nodule image based on the actual size corresponding to a single pixel of the ruler and the overall number of pixels of the target nodule image; 3) The average size of a single nodule is estimated by using the minimum circumscribed circle of image recognition, and the number parameter of the nodule is obtained by dividing the overall actual size by the average size of the single nodule.
