An aid for standardized acquisition of crop phenotyping and method of use

Abstract

The application discloses an auxiliary device for crop phenotype standardization collection and a use method. The device comprises a positioning background plate, a calibration function area, a sample fixing area, a shooting support, an information recording area and the like, adopts a modular design, can be spliced with a whole plant collection expansion board, and meets the requirements of crop local and whole plant image collection. Through the design of a matrix type fine hole with outer sparse and inner dense and outer small and inner large, the surface and the fluff of the crop are interacted to stably adsorb sample materials, combined with a standardized shooting background, a calibration reference and a sample posture adjustment scheme, an ordinary smart phone can collect a standardized image with high quality and which can be used for automatic trait extraction. The developed auxiliary algorithm can quickly process and extract data from the standardized image. Compared with the prior art, the application has the advantages of low cost, simple operation, good sample adaptability and high data comparability, and is suitable for crop phenotype rapid collection in a field and a laboratory environment.

Description

Technical Field

[0001] This invention belongs to the fields of smart agricultural equipment technology, computer vision technology, and biometric recognition technology, specifically relating to an auxiliary device and its usage method for standardized crop phenotypic data collection. Background Technology

[0002] Accurate and high-throughput acquisition of crop phenotypes, especially images, is currently a bottleneck in bio-breeding and a key to cultivation management. Traditional methods relying on manual measurement are no longer sufficient in terms of efficiency, objectivity, and data dimensionality in the era of big data phenotyping. Current crop phenotype technologies mainly suffer from the following gradient problems.

[0003] First, data acquisition is non-standardized. Researchers directly use mobile phones or cameras to take pictures in the field, lacking unified control, resulting in inconsistent image quality. Second, trait extraction is subjective and one-sided. Even when usable images are obtained, traditional methods can only manually measure a few simple traits such as "ear length" and "plant height," failing to quantify deeper image traits that reflect yield potential, stress resistance, and physiological state, such as texture entropy, fractal dimension, and color area ratio, which are involved in high-throughput bio-breeding screening. Finally, cost and accessibility are bottlenecks. Fully automated field phenotyping robots or large imaging platforms are limited by cost and space constraints, making widespread application in breeding bases and fields difficult. Therefore, the market urgently needs a compromise and efficient solution: one that provides a standardized data acquisition front-end, lays a solid foundation for subsequent automated, multi-dimensional trait extraction, and is low-cost, easy to operate, and portable, reducing the time, space, and labor costs of use and training.

[0004] Publication No. CN212624062U proposes an image acquisition auxiliary device. Its acquisition board integrates a background area, standard color blocks, a QR code, and a handwriting area. Image angles and colors are corrected using ruler frames and color blocks, and the QR code information is automatically labeled and archived. However, this device is not optimized for crop phenotypic acquisition, lacks a dedicated sample fixing mechanism, and cannot meet the standardized acquisition needs of specific parts such as panicles and whole plants. Publication No. CN104776800A proposes a rice single-husk image acquisition device and morphological parameter extraction method. Images are acquired using a black background board, a stem fixing arm, and calibration paper. Morphological parameters such as leaf angles are extracted through coordinate transformation and curve fitting. However, this device is only suitable for rice single-husk images, requires connection to a computer and specialized software, is difficult to deploy quickly in the field, and cannot simultaneously meet the phenotypic acquisition needs of panicles and whole plants. Publication No. CN110619297A proposes a batch acquisition and recognition method and device for legume fruit images. Its background plate has a placement slot array, identifies seeds and segments images using Hough circle detection, and automatically names and outputs images based on QR code information. However, this device is designed specifically for granular seeds and cannot be used for complex morphological samples such as ears or whole plants, and it lacks a posture control mechanism.

[0005] A comprehensive analysis of existing technologies reveals that traditional crop fixation and imaging devices mostly use mechanical clamping methods such as hairpins, clips, and clamps in conjunction with a background board to fix plant samples. However, this method involves concentrated clamping points, making the samples prone to shaking and susceptible to the effects of wind and water. Furthermore, single-point mechanical clamping often leads to twisting and folding of the sample tips, especially in the ears and leaves of gramineous crops, where rotation and overlap are more likely, causing errors in subsequent image processing and trait extraction. More importantly, for live samples collected for continuous trait acquisition, mechanical clamping may damage or abrade the sample surface, affecting normal growth. While some devices use multi-point clamping with straps, this is not only cumbersome to operate but also hinders subsequent processing. Although it can perform correction and imaging functions, it lacks adaptability to various plant types and samples. With the continuous advancement of optical imaging equipment, storage devices, and information technology, especially the development of deep learning technology based on big data, computers possess powerful image processing capabilities, placing higher demands on plant phenotypic trait extraction. Therefore, there is a greater need for a device suitable for various crop samples, easy to deploy and portable, capable of conveniently capturing shake-free images suitable for standardized extraction of plant phenotypic traits. Summary of the Invention

[0006] (a) Technical problems to be solved Traditional plant fixation and photography devices typically use iron hairpins, clips, and other similar devices in conjunction with a background board to secure and photograph plants. These devices are prone to shaking due to wind and water flow, resulting in plants that are not properly spread out in the photographs. Mechanical single-point clamping often leads to twisting and folding of the sample tips, causing damage and errors. Furthermore, they are unsuitable for plants of different sizes, making it difficult to balance operational complexity and sample fixation stability. Images corrected using simple correction methods are complex to process due to issues such as camera distortion and color region localization. While using large-scale equipment such as field robots ensures the accuracy and stability of the photographs, the long training time for operators, high equipment costs, and difficulties in deployment and use still limit their large-scale use.

[0007] (II) Technical Solution To address the aforementioned problems, this invention provides the following technical solution: an auxiliary device and method for standardized crop phenotypic data collection are proposed, as detailed below.

[0008] An auxiliary device for crop phenotypic standardization acquisition, which can assist in the rapid standardization of crop image phenotypic acquisition, includes a positioning background plate (1), a calibration function area (2), a sample fixing area (3), an imaging bracket (4), an information recording area (5), and an extension plate (6); wherein, The positioning background plate (1) is a matte black rigid material plate used to provide a high-contrast shooting background without reflection; the positioning background plate (1-1) is provided with a groove (1-1), and the surface of the plate is densely covered with fine holes extending obliquely inward, which are used to eliminate reflection and assist in fixing the plant sample. The calibration function area (2) is integrated on the positioning background plate (1) and includes a standard color card (2-1), a grayscale color card (2-2), a length scale (2-3) and a circular positioning calibration point (2-4), which are used for color correction, illumination normalization and geometric dimension calibration of the image; The sample fixing area (3) is set on the positioning background plate (1) and is used to fix the whole plant sample or organ sample to be tested in a predetermined posture to the strip-shaped slot (3-1) and groove (3-2) of the positioning background plate (1) and fix one sample at a time. The shooting bracket (4) is adjustable to the positioning background plate (1) via an adjustable connecting arm (4-1). The end is a universal mobile phone clip (4-2) for holding the shooting device and making the camera optical axis of the shooting device perpendicular to the positioning background plate (1). The camera field of view covers the sample fixing area (3) and the calibration function area (2). When shooting with smart glasses, the shooting bracket (4) can be used without using it or can be completely folded. The information recording area (5) is integrated on the positioning background plate (1) and includes a waterproof erasable handwriting plate (5-1) and a waterproof counter (5-2) for recording the variety information and shooting sequence number of the photographed sample, and for standardized information registration and reading in subsequent processing; The extension plate (6) can be optionally mounted on the groove (1-1) of the positioning background plate (1) and is the same size as the positioning background plate (1). It is used to fix and simultaneously capture images of the upper and lower parts of a larger sample for subsequent whole-plant morphology processing. It is equipped with the same standard color card (2-1), grayscale color card (2-2), length scale (2-3), and circular positioning calibration point (2-4) as the calibration function area (2).

[0009] Preferably, the fine holes on the surface of the positioning background plate (1) are designed as blind holes and are distributed in a matrix. The matrix shape is hexagonal or rectangular. The axis of the fine holes is designed to be inclined and forms an angle of 15-30 degrees with the normal of the plate surface.

[0010] Preferably, the pores on the surface of the positioning background plate (1) are non-uniformly distributed, and the pore density gradually decreases from the central region to the edge region; the pore size gradually decreases from the central region to the edge region.

[0011] Preferably, the positioning background plate (1) is coated with an antistatic coating and is sanded, with a surface roughness Ra value between 1.6 and 6.3 μm.

[0012] Preferably, the shooting device is a smartphone or smart glasses.

[0013] Preferably, the plant samples include any variety of rice, wheat, corn, buckwheat, oats, and barley.

[0014] This invention also includes a method of using an auxiliary device for crop phenotypic standardization acquisition, which uses the aforementioned auxiliary device for crop phenotypic standardization acquisition to acquire images, and the method of use includes the following steps: Step S1, fix the sample: open the strip-shaped bayonet (3-1) of the sample fixing area (3), adjust the sample into the embedded groove, install the expansion plate (6) as needed, and then close the bayonet; Step S2, Sample Registration: Register variety information on the waterproof erasable handwriting board (5-1) in the information recording area (5) by writing with a waterproof pen or sticking a barcode. Set the waterproof counter (5-2) to the starting position and mark the number of photos taken. Step S3: Bracket Adjustment and Imaging: Adjust the shooting bracket (4) to the preset position and take an image using the installed shooting device; Step S4: Image data processing: Process the acquired images, including information registration and extraction of crop root and aboveground part phenotypic data; Step S5: Repeat steps S1 to S4 until the entire batch of crop image acquisition and processing is completed, and the acquisition task ends. Step S6: Data Presentation: Construct a crop phenotype database corresponding to this round, establish a mapping relationship between crop phenotypes and crop images at each growth stage, and provide data for further research.

[0015] Preferably, the method in step S4 includes: Step S41, image fixed partition segmentation, that is, by defining the area width in advance, the information of the information recording area (5) is fixedly identified, and the sample variety information and serial number are obtained by using OCR or QR code recognition method; Step S42, actual crop image segmentation: the original plant image acquired in step S1 is segmented by the crop semantic segmentation model. After segmentation, the model sets the crop plant area to white and other areas to black to obtain a crop binary image. Step S43, image masking: mask the crop binary image and the original image to obtain independent segmented images of the crop plants; Step S44, image morphology acquisition, performing morphology processing on the crop independently segmented image based on machine vision algorithms, including using image convex hull extraction, texture morphology extraction, area pixel calculation, and grayscale value calculation to obtain the target morphology.

[0016] Preferably, in step S42, the crop semantic segmentation model segmentation specifically adopts a crop semantic segmentation model trained by the Segformer semantic segmentation network, which is obtained by manual annotation in advance. The model is used to segment multiple parts of the aboveground parts of the crop, including the main stem, tillers, ears, and leaves, and to remove impurities and perform image preprocessing.

[0017] Preferably, the traits extracted in step S44 include texture mean, texture smoothness, texture standard deviation, texture third-order distance, texture consistency, texture entropy, whole-plant projected area, height, width, ratio of whole-plant projected area to bounding rectangle area, whole-plant minimum convex hull area, ratio of total area to minimum convex hull area, perimeter-area ratio, fractal dimension, green projected area, yellow projected area, H-component gray-level co-occurrence matrix features, perimeter, and yellow projected area ratio.

[0018] (III) Beneficial Effects Compared with the prior art, the present invention has at least the following positive technical effects.

[0019] (1) Significantly improved sample fixation stability and adaptability. Traditional devices use hard mechanical clamping methods such as hairpins and buckles, which not only easily cause plant shaking, but also have difficulty adapting to samples of different sizes and shapes, and cannot guarantee the natural and relaxed posture of the sample during the shooting process. This invention sets a sample fixing area on the positioning background plate. Its strip-shaped slots and grooves can accurately position and flexibly clamp the plant stems. The magnetic strip-shaped slots facilitate quick replacement of crop samples and secure fixation. At the same time, the oblique fine holes densely distributed on the surface of the plate assist in fixing the sample. With a single strip-shaped slot and groove, the ears or whole plants of crops such as rice, wheat, barley, oats, and buckwheat can be stably attached to the background plate in a predetermined posture. In addition, the optional extension plate is spliced ​​with the positioning background plate through a sliding groove to provide a continuous background and fixation support for large whole plant samples such as corn, solving the problem that traditional devices cannot accommodate samples of different sizes.

[0020] (2) The oblique micropore design achieves the technical effect of assisting in sample fixation. The distributed micro-adsorption and friction of the micropores achieve non-destructive fixation. Unlike the point contact of traditional clamps, the micropore design uses several matrix-distributed micro-adsorption points to disperse the fixing force across the entire contact area between the sample and the plate surface. For the relatively heavier stem area, denser and coarser micropores are designed to enhance the fixing effect of friction. When a moist or finely hairy crop sample is placed on the plate, the micropores naturally "hang" the sample using capillary action, friction, and the micro-negative pressure effect generated by water, achieving the effect of "adsorption upon placement". This distributed adsorption method offers several advantages: it secures samples without additional clamping force, particularly stabilizing the leaf tips and significantly reducing mechanical damage; the adaptive distribution of adsorption force across different areas ensures balanced sample stress, preventing movement and allowing the entire sample to naturally and comfortably conform to the plate surface, greatly improving posture standardization; the pore axis forms a 15-30 degree angle with the plate surface normal, achieving directional adsorption, while the angled pore walls provide tangential friction, effectively resisting sample slippage. Furthermore, the frosted surface and pores work together to create a microtexture, further enhancing friction; the blind-pore design facilitates natural dust removal or rainwater washing, resulting in lower maintenance costs.

[0021] (3) The standardization, calibration accuracy, and portability of image acquisition are greatly improved. Existing simple calibration methods lack a unified reference and shooting conditions, making images susceptible to camera distortion, lighting changes, and color deviations, resulting in complex post-processing and significant errors. This invention integrates a calibration function area containing a standard color chart, grayscale chart, length scale, and circular positioning calibration points, providing an absolute benchmark for color, grayscale, and geometric dimensions for each image. All parts are mutually compatible, ensuring stable color, grayscale, and geometric dimension calibration even with partial damage or defects. Simultaneously, the shooting bracket can adjust and lock the shooting equipment, ensuring the camera's optical axis is strictly perpendicular to the positioning background plate, thus eliminating image standardization problems caused by random changes in shooting angle and distance. These features collectively ensure consistent image color reproduction, controllable geometric distortion, and accurate dimension calibration, making images acquired by different personnel, equipment, and environments highly comparable. Furthermore, this invention has a simple structure, is lightweight, and is compatible with ordinary smartphones or smart glasses as shooting devices. It requires no additional investment in professional imaging systems, is extremely easy to learn, and is easy to promote quickly and in large quantities. It is suitable for widespread application in breeding bases and field sites.

[0022] (4) Significantly expanded data processing efficiency and trait extraction dimensions. Traditional methods can only manually measure a few simple traits such as ear length and plant height. However, the image processing method of this invention, based on standardized images obtained using the device, automatically identifies different parts of the crop using a semantic segmentation model trained by Segformer, and can extract more than twenty deep phenotypic traits (such as texture entropy, fractal dimension, color area ratio, etc.) in batches, realizing a qualitative change from "images" to "structured data". In particular, the precise correction provided by the calibration function area makes the extraction of advanced traits such as texture features and color components repeatable and reliable.

[0023] (5) Optimization of environmental adaptability and durability. Traditional devices are prone to reflection, static electricity adsorption, or surface contamination in strong light, dust, or humid environments in the field, affecting imaging quality. The positioning background plate of this invention is made of matte black material, with a frosted surface and an antistatic coating, effectively eliminating reflection and static interference.

[0024] In summary, this invention, through the organic combination of modular hardware structure and intelligent image processing methods, solves the core problems in existing technologies such as unstable sample fixation, inconsistent image quality, high cost, and one-sided trait extraction. It achieves standardized, low-cost, and high-throughput crop phenotypic data collection, providing a practical and effective technical tool for agricultural research and breeding practices. Attached Figure Description

[0025] Figure 1 This is a detailed layout diagram of the positioning background plate of the present invention.

[0026] Figure 2 It is a schematic diagram of the strip卡口 and groove of the sample fixing area of the present invention.

[0027] Figure 3 It is a schematic diagram of the adsorption and friction of the positioning background plate of the present invention.

[0028] Figure 4 It is a schematic diagram of the working state of the shooting bracket of the present invention.

[0029] Figure 5 It is a schematic diagram of the detailed layout of the extension board of the present invention.

[0030] Figure 6 It is a pipeline flow chart of data collection and processing of the present invention.

[0031] Figure 7 It is a pipeline diagram of the collection program processing supporting the present invention. Specific implementation manners

[0032] The present invention provides an auxiliary device and a usage method for standardized collection of crop phenotypes. The main implementation steps include an image shooting device and its assembly; collection, processing, and data extraction of crop sample traits. The present invention will be further described below in conjunction with the accompanying drawings and implementation cases.

[0033] (I) Image shooting device and its assembly As Figure 1 shown, an auxiliary device for standardized collection of crop phenotypes provided by the present invention mainly includes a positioning background plate (1), a calibration function area (2), a sample fixing area (3), a shooting bracket (4), and an information recording area (5). Among them, the positioning background plate (1) is the basic bearing component of the entire device. It is made of a matte black rigid material plate, preferably acrylic or ABS engineering plastic to meet the requirements of both lightness and strength, and can accommodate crop samples of sufficient size. The surface of this plate is treated by sanding, and the roughness Ra value is controlled between 1.6 - 6.3 μm, and an antistatic coating is applied to eliminate reflection and static interference.

[0034] As Figure 1As shown, the calibration function area (2) is mainly integrated in the middle and lower left area of ​​the positioning background plate (1), including a standard color card (2-1), a grayscale color card (2-2), a length scale (2-3), and circular positioning calibration points (2-4). The standard color card (2-1) uses internationally recognized 18-color standard color blocks for image color correction; the grayscale color card (2-2) contains a 6-level 8-bit grayscale scale from white to black for illumination normalization and grayscale color calibration, and the uniform color block size allows it to also serve as a geometric dimension auxiliary calibration function; the length scale (2-3) is a precision scale with millimeters as the smallest unit for geometric dimension calibration; the circular positioning calibration point (2-4) consists of multiple high-contrast dots for image distortion correction and viewing angle calibration, and its fixed size can also assist in dimension calibration. All calibration elements are printed using a waterproof process to ensure long-term use in field environments.

[0035] like Figure 1 As shown, the information recording area (5) is integrated above the positioning background plate (1), including a waterproof erasable writing board (5-1) and a waterproof counter (5-2). The waterproof erasable writing board (5-1) is made of white matte material and can be used to write information such as sample type and date with a waterproof pen. The written content can be erased and reused, and printed barcodes or QR codes can also be pasted on it. The waterproof counter (5-2) is a mechanical or electronic counter used to record the shooting sequence number. The design of this area allows the sample information to be directly integrated into the shooting image, which is convenient for subsequent automated recognition and archiving.

[0036] like Figure 2 As shown, the sample fixing area (3) is located in the lower middle part of the positioning background plate (1), including a strip-shaped latch (3-1) and a groove (3-2). The strip-shaped latch (3-1) is made of stainless steel and is magnetic, used to cooperate with the corresponding position of the background plate to press the sample stem; the groove (3-2) is an arc-shaped groove that matches the stem, used for positioning and limiting. In use, first remove the strip-shaped latch (3-1), insert the stem part of the sample to be tested into the groove (3-2) from top to bottom, then align the strip-shaped latch back to the appropriate position, and after adsorption, the sample can be fixed on the background plate in a predetermined posture (such as natural side extension). For whole plant samples, the main stem and tillers can be fixed by adding multiple latches and grooves in combination.

[0037] like Figure 3As shown, the positioning background plate (1) has a dense network of fine pores extending obliquely inward on its surface. These pores are designed as blind pores and are distributed in a matrix (preferably hexagonal or rectangular array). The axis of the pores forms an angle of 15-30 degrees with the normal to the plate surface, creating a unidirectional micro-adsorption force. When a wet or slightly burred plant sample is placed on the plate, the pores can naturally "hold" the sample by relying on capillary action, friction, and the micro-negative pressure effect generated by water, preventing displacement caused by a breeze or operational disturbance. In addition, the pores are non-uniformly distributed, with a higher density in the central area and a lower density in the edge area, and the pore size decreases accordingly. This ensures the adsorption effect in the core fixation area while allowing the edge area to obtain sufficient adsorption force with reduced damage.

[0038] like Figure 4 As shown, the shooting bracket (4) is connected to the top edge of the positioning background plate (1) via an adjustable connecting arm (4-1). The connecting arm (4-1) adopts a foldable and telescopic structure, which can preset different shooting heights and angles. The end of the shooting bracket (4) is a universal mobile phone clip (4-2), which can hold smartphones of different sizes. If using smart glasses for shooting, the shooting bracket (4) can be omitted or completely folded.

[0039] like Figure 5 As shown, the extension plate (6) can be optionally mounted on the groove (1-1) of the positioning background plate (1), and is the same size as the positioning background plate (1). It is used to fix the sample and simultaneously capture images of the upper and lower parts of the larger sample for subsequent whole-plant morphology processing. The extension plate (6) is equipped with the same standard color card (2-1), grayscale color card (2-2), length scale (2-3), and circular positioning calibration point (2-4) as the calibration functional area (2), ensuring that the whole plant is still equipped with a complete calibration reference. The surface of the extension plate is also densely covered with oblique fine holes for fixing the extended part of the plant.

[0040] After the shooting bracket (4) is stably connected to the positioning background plate (1) and the expansion plate (6) is installed according to the task requirements, the shooting equipment is installed and then the image can be captured.

[0041] (II) Crop Sample Trait Collection, Processing, and Data Extraction Combination Figure 6 and Figure 7 The method of using the auxiliary device of the present invention includes the following steps: Step S1, fix the sample: open the strip-shaped bayonet (3-1) of the sample fixing area (3), adjust the sample into the embedded groove, install the expansion plate (6) as needed, and then close the bayonet; Step S2, Sample Registration: Register variety information on the waterproof erasable handwriting board (5-1) in the information recording area (5) by writing with a waterproof pen or sticking a barcode. Set the waterproof counter (5-2) to the starting position and mark the number of photos taken. Step S3: Bracket Adjustment and Imaging: Adjust the shooting bracket (4) to the preset position and take an image using the installed shooting device; Step S4: Image data processing: Process the acquired images, including information registration and extraction of crop root and aboveground part phenotypic data; Step S5: Repeat steps S1 to S4 until the entire batch of crop image acquisition and processing is completed, and the acquisition task ends. Step S6: Data Presentation: Construct a crop phenotype database corresponding to this round, establish a mapping relationship between crop phenotypes and crop images at each growth stage, and provide data for further research.

[0042] Preferably, the method in step S4 includes: Step S41, image fixed partition segmentation, that is, by defining the area width in advance, the information of the information recording area (5) is fixedly identified, and the sample variety information and serial number are obtained by using OCR or QR code recognition method; Step S42, actual crop image segmentation: the original plant image acquired in step S1 is segmented by the crop semantic segmentation model. After segmentation, the model sets the crop plant area to white and other areas to black to obtain a crop binary image. Step S43, image masking: mask the crop binary image and the original image to obtain independent segmented images of the crop plants; Step S44, image morphology acquisition, performing morphology processing on the crop independently segmented image based on machine vision algorithms, including using image convex hull extraction, texture morphology extraction, area pixel calculation, and grayscale value calculation to obtain the target morphology.

[0043] This invention achieves a fundamental transformation in crop phenotyping from "casual photography" to "standardized data production" through the deep integration of physical mechanisms and intelligent algorithms. At the hardware level, the innovative oblique micropore array utilizes the synergistic effect of capillary adsorption, mechanical oblique friction, and micro-negative pressure to completely solve the industry pain points of traditional clamps damaging samples and cumbersome operations, achieving non-destructive and efficient collection through "placement and fixation." The integrated calibration function area and modular adaptive expansion design ensure that anyone can obtain high-quality images with color uniformity, geometric calibration, and consistent perspective in a field environment. At the software level, the accompanying intelligent analysis system, based on the standardized images collected by the device, can batch extract more than twenty deep phenotypic traits in four categories: texture, morphology, color, and complexity, directly converting images into structured data. The entire system upgrades ordinary smartphones into professional phenotyping tools at extremely low cost, significantly outperforming existing technologies in three core indicators: fixation stability, data standardization, and trait extraction dimensions. It provides a high-throughput solution for breeding research and precision agriculture that is "plug and play in the field, data is obtained instantly upon capture."

[0044] The specific examples described in this application are merely illustrative of the spirit of the invention. Those skilled in the art to which this invention pertains may make various modifications or additions to the specific examples described herein, or substitute them by similar means, without departing from the spirit of the invention or exceeding the scope defined by the appended claims.

Claims

1. An auxiliary device for crop phenotypic standardization acquisition, which can assist in the rapid standardization of crop image phenotypic acquisition, characterized in that, The device includes a positioning background plate (1), a calibration function area (2), a sample fixing area (3), a shooting bracket (4), an information recording area (5), and an expansion plate (6); among which, The positioning background plate (1) is a matte black rigid material plate used to provide a high-contrast shooting background without reflection; the positioning background plate (1-1) is provided with a groove (1-1), and the surface of the plate is densely covered with fine holes extending obliquely inward, which are used to eliminate reflection and assist in fixing the plant sample. The calibration function area (2) is integrated on the positioning background plate (1) and includes a standard color card (2-1), a grayscale color card (2-2), a length scale (2-3) and a circular positioning calibration point (2-4), which are used for color correction, illumination normalization and geometric dimension calibration of the image; The sample fixing area (3) is set on the positioning background plate (1) and is used to fix the whole plant sample or organ sample to be tested in a predetermined posture to the strip-shaped slot (3-1) and groove (3-2) of the positioning background plate (1) and fix one sample at a time. The shooting bracket (4) is adjustablely connected to the positioning background plate (1) via an adjustable connecting arm (4-1). The end is a universal mobile phone clip (4-2) for holding the shooting device and making the camera optical axis of the shooting device perpendicular to the positioning background plate (1), and making the camera field of view cover the sample fixing area (3) and the calibration function area (2). The information recording area (5) is integrated on the positioning background plate (1) and includes a waterproof erasable handwriting plate (5-1) and a waterproof counter (5-2) for recording the variety information and shooting sequence number of the photographed sample, and for standardized information registration and reading in subsequent processing; The extension plate (6) can be optionally mounted on the groove (1-1) of the positioning background plate (1) and is the same size as the positioning background plate (1). It is used to fix and simultaneously capture images of the upper and lower parts of a larger sample for subsequent whole-plant morphology processing. It is equipped with the same standard color card (2-1), grayscale color card (2-2), length scale (2-3), and circular positioning calibration point (2-4) as the calibration function area (2).

2. The auxiliary device for standardized crop phenotypic data collection according to claim 1, characterized in that, The fine holes on the surface of the positioning background plate (1) are designed as blind holes and are distributed in a matrix. The matrix shape is hexagonal or rectangular. The axis of the fine holes is designed to be inclined and forms an angle of 15-30 degrees with the normal of the plate surface.

3. The auxiliary device for standardized crop phenotypic data collection according to claim 1, characterized in that, The pores on the surface of the positioning background plate (1) are non-uniformly distributed, and the density of the pores gradually decreases from the central region to the edge region; the size of the pores gradually decreases from the central region to the edge region.

4. The auxiliary device for crop phenotypic standardization data collection according to claim 1, characterized in that, The positioning background plate (1) is coated with an antistatic coating and is sanded, with a surface roughness Ra value between 1.6 and 6.3 μm.

5. The auxiliary device for standardized crop phenotypic data collection according to claim 1, characterized in that, The shooting device is a smartphone or smart glasses.

6. The auxiliary device for standardized crop phenotypic data collection according to claim 1, characterized in that, The plant samples include any variety of rice, wheat, corn, buckwheat, oats, and barley.

7. A method of using an auxiliary device for crop phenotypic standardization acquisition, wherein the auxiliary device for crop phenotypic standardization acquisition as described in any one of claims 1-6 is used for image acquisition, and the method of use includes the following steps: Step S1, fix the sample: open the strip-shaped bayonet (3-1) of the sample fixing area (3), adjust the sample into the embedded groove, install the expansion plate (6) as needed, and then close the bayonet; Step S2, Sample Registration: Register variety information on the waterproof erasable handwriting board (5-1) in the information recording area (5) by writing with a waterproof pen or pasting a barcode. Set the waterproof counter (5-2) to the starting position and mark the number of photos taken. Step S3: Bracket Adjustment and Imaging: Adjust the shooting bracket (4) to the preset position and take an image using the installed shooting device; Step S4: Image data processing: Process the acquired images, including information registration and extraction of crop root and aboveground part phenotypic data; Step S5: Repeat steps S1 to S4 until the entire batch of crop image acquisition and processing is completed, and the acquisition task ends. Step S6: Data Presentation: Construct a crop phenotype database corresponding to this round, establish a mapping relationship between crop phenotypes and crop images at each growth stage, and provide data for further research.

8. The method of using the auxiliary device for crop phenotypic standardization data collection according to claim 7, characterized in that, The method in step S4 includes: Step S41, image fixed partition segmentation, that is, by defining the area width in advance, the information of the information recording area (5) is fixedly identified, and the sample variety information and serial number are obtained by using OCR or QR code recognition method; Step S42, actual crop image segmentation: the original plant image acquired in step S1 is segmented by the crop semantic segmentation model. After segmentation, the model sets the crop plant area to white and other areas to black to obtain a crop binary image. Step S43, image masking: mask the crop binary image and the original image to obtain independent segmented images of the crop plants; Step S44, image morphology acquisition, performing morphology processing on the crop independently segmented image based on machine vision algorithms, including using image convex hull extraction, texture morphology extraction, area pixel calculation, and grayscale value calculation to obtain the target morphology.

9. The method of using the auxiliary device for crop phenotypic standardization data collection according to claim 8, characterized in that, In step S42, the crop semantic segmentation model is specifically used to segment the crop semantic segmentation model trained by the Segformer semantic segmentation network, which is obtained by manual annotation in advance. The model is used to segment multiple parts of the aboveground parts of the crop, including the main stem, tillers, ears and leaves, and to remove impurities and perform image preprocessing.

10. The method of using the auxiliary device for crop phenotypic standardization data collection according to claim 8, characterized in that, The traits extracted in step S44 include texture mean, texture smoothness, texture standard deviation, texture third distance, texture consistency, texture entropy, whole plant projected area, height, width, ratio of whole plant projected area to bounding rectangle area, whole plant minimum convex hull area, ratio of total area to minimum convex hull area, perimeter-area ratio, fractal dimension, green projected area, yellow projected area, H component gray-level co-occurrence matrix features, perimeter, and yellow projected area ratio.

Images