A method and system for measuring the phenotype of a field corn plant

Abstract

A field corn plant phenotype measurement method and system, the method first uses an image collector to shoot an RGB image and a depth image of a field corn plant, then extracts a mask image from the depth image based on a clip module added in a YOLOv8 model, and performs mask processing on the RGB image, then uses the YOLOv8 model to perform target recognition on the RGB image after mask processing, then determines the key point pixel coordinates output by the recognized target, then extracts the coordinate depth corresponding to the key point pixel coordinates from the depth image, converts the camera coordinate system coordinates of the key points into world coordinate system coordinates after calculating, determines the key points belonging to the same plant through coordinate matching and pairing, and finally calculates the ear height and leaf angle of the corn plant according to the world coordinate system coordinates of the paired key points. The present application can realize accurate analysis of the ear height and leaf angle phenotype of low-resolution images.

Description

Technical Field

[0001] This invention belongs to the field of agricultural bioinformatics technology, specifically relating to a method and system for measuring the phenotypic characteristics of field maize plants. Background Technology

[0002] Crop phenotypic information acquisition is a crucial step in modern breeding technology. With the development of gene sequencing technology, the vast array of gene sequences has generated a massive number of crop phenotypic samples. By analyzing the traits of these samples, breeders can screen for high-quality, high-yield crop varieties, thereby meeting the rapidly growing global food demand.

[0003] Currently, the measurement of maize plant phenotypes in agriculture still relies primarily on manual methods. Published literature mainly utilizes binocular RGBD cameras and robotic vehicles for image-based phenotypic analysis. This method not only involves complex equipment but also suffers from limitations due to the narrow field of view of RGBD cameras, making it difficult to obtain comprehensive information from root to ear in a single image. It requires a relatively large shooting distance or a large tilt angle, or even multiple cameras to meet the viewing angle requirements. However, the row spacing in open fields does not allow for large shooting distances, and large tilt angles severely affect the accuracy of depth values. High-resolution depth maps still suffer from severe compression in areas far from the lens, with the compression intensifying at larger tilt angles. Therefore, this method is largely limited to potted plant scenarios and is unsuitable for open fields. Summary of the Invention

[0004] The purpose of this invention is to address the aforementioned problems in the existing technology by providing a simple and highly accurate method and system for measuring the phenotypic characteristics of field maize plants.

[0005] To achieve the above objectives, the technical solution of the present invention is as follows:

[0006] In a first aspect, the present invention proposes a method for measuring the phenotypic characteristics of maize plants in the field, comprising:

[0007] The method includes:

[0008] S1. Use an image acquisition device to capture RGB images and depth images of maize plants in the field. The RGB images and depth images both include close-up and distant images.

[0009] S2. Based on the clip module added to the YOLOv8 model, extract the mask image from the depth image and apply the extracted mask image to the RGB image. Extract the RGB pixels covered by the mask image as the RGB image after mask processing.

[0010] S3. Use the YOLOv8 model to perform target recognition on the masked RGB image. For distant images, the target is identified as corn aerial roots and the segmented target is corn female ears. For close-up images, the target is identified as corn leaf nodes.

[0011] S4. Determine the key pixel coordinates of the identified target. The key pixel coordinates of the corn aerial roots are the coordinates of the points in the target box. The key pixel coordinates of the corn female ear are the coordinates of the lowest point of the segmented region. The key pixel coordinates of the corn leaf node are the coordinates of point S on the leaf vein that is less than 50 pixels away from the leaf node, point M at the junction of the leaf and the stem, and point C on the stem that is less than 50 pixels away from the leaf node.

[0012] S5. Extract the coordinate depth D corresponding to the pixel coordinates of the key point from the depth image, and calculate the camera coordinate system coordinates of the key point.

[0013] S6. Based on the camera coordinate system coordinates of the key points and the posture data of the image acquisition device, convert their coordinates to world coordinate system coordinates, determine the key points belonging to the same plant through coordinate matching, and pair them up.

[0014] S7. Calculate the ear height and leaf angle of the maize plant based on the world coordinate system coordinates of the key points of pairing.

[0015] The clip module is located before the backbone module of YOLOv8;

[0016] In step S2, the extraction of the mask image from the depth image by the clip module added to the YOLOv8 model includes:

[0017] The clip module first determines the cut-off depth range based on the shooting distance, and then extracts the pixels within the cut-off depth range from the depth image as a mask image. The cut-off depth range is [d...]. min ,d min +0.3(d max -d min )],d max d min These are the maximum and minimum depth values ​​that can be read from the depth image, respectively.

[0018] In step S5, while extracting the coordinate depth D corresponding to the pixel coordinates of the key point from the depth image, the depth confidence of the point is determined. If the depth confidence is lower than the threshold, the depth D is replaced by the average depth of multiple pixels around the point with confidence values ​​higher than the threshold. If the confidence values ​​of all pixels around the point are lower than the threshold, the depth image is discarded.

[0019] In S7, the ear height is calculated according to the following formula:

[0020]

[0021] In the above formula, EH represents the ear height. The three-dimensional coordinates of the key points of the female ear of maize. Three-dimensional coordinates of key points for corn aerial roots;

[0022] The blade angle is calculated using the following formula:

[0023]

[0024] In the above formula, LA is the leaf angle, SM is the distance between points S and M, MC is the distance between points M and S, and SC is the distance between points S and C. These are the three-dimensional coordinates of points S, M, and C, respectively.

[0025] The image acquisition device is a mobile phone equipped with a TOF sensor. When shooting distant images, the distance between the image acquisition device and the corn plant is not less than 0.75 meters; when shooting close-up images, the distance between the image acquisition device and the corn plant is 0.2 meters; the pitch angle and roll angle of the image acquisition device are both within (-30°, 30°).

[0026] S6 determines key points belonging to the same plant based on the Z and X value ranges of the world coordinate system coordinates of the key points, wherein the Z and X value ranges are both less than 0.15 meters.

[0027] Secondly, this invention proposes a field maize plant phenotypic measurement system, including an image acquisition device, a YOLOv8 model, a key point pixel coordinate determination module, a key point camera coordinate determination module, a key point pairing module, and a module for calculating ear height and leaf angle.

[0028] The image acquisition device is used to capture RGB images and depth images of corn plants in the field. Both the RGB images and depth images include close-up and distant images.

[0029] The YOLOv8 model is used for:

[0030] The added clip module extracts a mask image from the depth image and applies the extracted mask image to the RGB image, extracting the RGB pixels covered by the mask image as the RGB image after masking.

[0031] Target recognition is performed on the RGB image after masking. For distant images, the target is identified as corn aerial roots and the segmented target is corn female ears. For close-up images, the target is identified as corn leaf nodes.

[0032] The key point pixel coordinate determination module is used to determine the key point pixel coordinates output by the identified target. Among them, the key point pixel coordinates of the corn aerial roots are the coordinates of the center point of the target box, the key point pixel coordinates of the corn female ear are the coordinates of the lowest point of the segmented area, and the key point pixel coordinates of the corn leaf node are the coordinates of point S on the leaf vein that is less than 50 pixels away from the leaf node, point M where the leaf and stem meet, and point C on the stem that is less than 50 pixels away from the leaf node.

[0033] The key point camera coordinate determination module is used to extract the coordinate depth D corresponding to the pixel coordinates of the key point from the depth image, and calculate the camera coordinate system coordinates of the key point by combining the pose and latitude and longitude data of the image acquisition device.

[0034] The key point pairing module is used to convert the coordinates of the key points into world coordinates based on the camera coordinate system coordinates and the posture data of the image acquisition device, and to determine and pair key points belonging to the same plant through coordinate matching.

[0035] The ear height and leaf angle calculation module is used to calculate the ear height and leaf angle of the maize plant based on the world coordinate system coordinates of the paired key points.

[0036] The clip module is located before the backbone module of YOLOv8;

[0037] The extraction of the mask image from the depth image includes:

[0038] The clip module first determines the cut-off depth range based on the shooting distance, and then extracts the pixels within the cut-off depth range from the depth image as a mask image. The cut-off depth range is [d...]. min ,d min +0.3(d max -d min )],d max d min These are the maximum and minimum depth values ​​that can be read from the depth image, respectively.

[0039] The key point camera coordinate determination module extracts the coordinate depth D corresponding to the key point pixel coordinates from the depth image and determines the depth confidence of the point. If the depth confidence is lower than the threshold, the depth D is replaced by the average depth of multiple pixels around the point with confidence higher than the threshold. If the confidence of all pixels around the point is lower than the threshold, the depth image is discarded.

[0040] The ear height is calculated using the following formula:

[0041]

[0042] In the above formula, EH represents the ear height. The three-dimensional coordinates of the key points of the female ear of maize. Three-dimensional coordinates of key points for corn aerial roots;

[0043] The blade angle is calculated using the following formula:

[0044]

[0045] In the above formula, LA is the leaf angle, SM is the distance between points S and M, MC is the distance between points M and S, and SC is the distance between points S and C. These are the three-dimensional coordinates of points S, M, and C, respectively.

[0046] The image acquisition device is a mobile phone equipped with a TOF sensor. When shooting distant images, the distance between the image acquisition device and the corn plant is not less than 0.75 meters; when shooting close-up images, the distance between the image acquisition device and the corn plant is 0.2 meters; the pitch angle and roll angle of the image acquisition device are both within (-30°, 30°).

[0047] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0048] 1. The present invention provides a method for measuring the phenotypic characteristics of field maize plants. First, an image acquisition device is used to capture RGB and depth images of field maize plants. Then, based on the clip module added to the YOLOv8 model, a mask image is extracted from the depth image. The extracted mask image is applied to the RGB image, and the RGB pixels covered by the mask image are extracted as the masked RGB image. Then, the YOLOv8 model is used to perform target recognition on the masked RGB image. Next, the key point pixel coordinates output by the identified target are determined. Then, the coordinate depth D corresponding to the key point pixel coordinates is extracted from the depth image, and the camera coordinate system coordinates of the key points are calculated. Finally, the coordinates of the key points are converted to world coordinate system coordinates based on the camera coordinate system coordinates of the key points and the pose data of the image acquisition device. Key points belonging to the same plant are determined by coordinate matching and paired. Based on the world coordinate system coordinates of the paired key points, the ear height and leaf angle of the maize plant are calculated. On the one hand, this method can analyze phenotypes such as ear height and leaf angle simultaneously for very low-resolution depth maps (such as depth maps on mobile phones, with a resolution of only 256*192), with accurate analysis results and a relative error of less than 5%. On the other hand, this method has a simple process, does not require the establishment of point clouds, has low requirements for computing resources, can perform edge computing, can be directly deployed on mobile devices (such as mobile phones), and perform real-time measurements.

[0049] 2. The present invention provides a method for measuring the phenotypic characteristics of maize plants in the field. This method extracts the coordinate depth D corresponding to the pixel coordinates of the key point from a depth image and simultaneously determines the depth confidence of that point. If the depth confidence is lower than a threshold, the depth D is replaced by the average depth of multiple pixels around that point with confidence levels higher than the threshold. If the confidence levels of all pixels around that point are lower than the threshold, the depth image is discarded. This strategy effectively improves the accuracy of the phenotypic measurement results. Furthermore, this method utilizes the clip module added to the YOLOv8 model to extract a mask image from the depth image based on the shooting distance, eliminating background interference and making the key point recognition highly accurate, with a map50 accuracy exceeding 0.93, thereby further ensuring the accuracy of the phenotypic measurement results. Attached Figure Description

[0050] Figure 1 This is a flowchart of the method described in Example 1.

[0051] Figure 2 This is a schematic diagram of the structure of the YOLOv8 model in Example 1.

[0052] Figure 3 The recognition result is shown in Example 1 for the close-up image.

[0053] Figure 4 The image shown is the recognition result of the distant view in Example 1.

[0054] Figure 5 This is a structural diagram of the system described in Example 2. Detailed Implementation

[0055] The present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0056] This invention proposes a method for measuring the phenotypic characteristics of maize plants in the field. This method is based on a monocular camera and a Time-of-Flight (TOF) sensor, and is applicable to any mobile phone equipped with a TOF sensor, with very low resolution requirements for the depth camera. Furthermore, the correlation index R of the ear height measurement results obtained by this method is [not specified in the original text]. 2 The correlation index R of the leaf angle results was 0.81, RMSE was 4 cm, MAE was 3 cm, and the correlation index R of the leaf angle results was 4 cm. 2 The value was 0.798, the RMSE was 6.8°, and the MAE was 4.4°.

[0057] Example 1:

[0058] A method for measuring the phenotypic characteristics of maize plants in the field, such as Figure 1 As shown, the specific steps are as follows:

[0059] 1. A mobile phone equipped with a TOF sensor was used to collect data on corn plants in the field. The phone's main camera captured RGB images, including close-up and distant images. The TOF sensor captured the required depth images, also including close-up and distant images. During image capture, the phone's current roll angle relative to the X-axis, pitch angle relative to the Y-axis, and yaw angle relative to the Z-axis were recorded, along with the phone's current latitude, longitude, and elevation coordinates recorded by an RTK device. When capturing distant images, the distance between the image acquisition device and the corn plant was no less than 0.75 meters; when capturing close-up images, the distance was 0.2 meters; the pitch and roll angles of the image acquisition device were limited to within (-30°, 30°).

[0060] 2. The acquired depth image is filled using the nearest neighbor interpolation method to make its resolution the same as that of the RGB image.

[0061] 3. The YOLOv8 model includes a clip module, which precedes the backbone module. Both the RGB and depth images serve as input to YOLOv8. The clip module then truncates and crops the RGB image before using it as input for the entire detection process. Figure 2 As shown.

[0062] The truncation and trimming include:

[0063] The clip module first determines the cut-off depth range based on the shooting distance, and then extracts pixels within the cut-off depth range from the depth image as a mask image. The cut-off depth range is [d...]. min ,d min +0.3(d max -d min )],d max d min These are the maximum and minimum depth values ​​that can be read from the depth image, respectively.

[0064] 4. Apply the mask image to the RGB image, and through image overlay analysis, extract the RGB pixels covered by the mask image as the RGB image after masking.

[0065] 5. The YOLOv8 model was used to perform target recognition on the masked RGB images. For distant images, the target was identified as corn aerial roots and the segmented target was corn female ears. For close-up images, the target was identified as corn leaf nodes.

[0066] The recognition results for distant images and distant images are as follows: Figure 3 , Figure 4 As shown.

[0067] 6. Determine the key pixel coordinates of the identified targets. The key pixel coordinates of the corn aerial roots are the coordinates of the center point of the target box. The key pixel coordinates of the corn female ear are the coordinates of the lowest point of the segmented region. The key pixel coordinates of the corn leaf nodes are the coordinates of point S on the leaf vein that is less than 50 pixels away from the leaf node, point M at the junction of the leaf and the stem, and point C on the stem that is less than 50 pixels away from the leaf node.

[0068] The key pixel coordinates of the near and far views are shown in Table 1 and Table 2, respectively:

[0069] Table 1 Key point pixel coordinates of close-up images

[0070] key_point Mx key_point Mx key_point Cx 672 672 693 key_point Cx key_point Sx key_point Sy 693 694 720

[0071] Table 2 Key pixel coordinates of distant view images

[0072] ear_x ear_y root_x root_y 740 928 673 1869

[0073] 7. Extract the coordinate depth D corresponding to the pixel coordinates of the key point from the depth image, and determine the depth confidence of the point. If the depth confidence is lower than the threshold, replace the depth D with the average depth of the 8 pixels around the point with confidence higher than the threshold. If the confidence of all pixels around the point is lower than the threshold, discard the depth image.

[0074] 8. Convert the pixel coordinates of key points to world coordinates.

[0075] Because there is a difference between the pixel coordinates of an image and the actual spatial coordinates, a coordinate system transformation is required. The transformation involves three processes: pixel coordinate system → image coordinate system → camera coordinate system → world coordinate system.

[0076] The transformation from pixel coordinates to image coordinates is performed according to the following formula:

[0077] P image (u, v) = P pixel (x, y)-P center (C x C y )

[0078] In the above formula, P pixel (x, y) are the coordinates of the pixel, P center (C x C y P represents the coordinates of the image center. image (u, v) are the coordinates of the transformed image;

[0079] The transformation from image coordinates to world coordinates is achieved through the rotation matrix R, the translation vector t, and the intrinsic parameter matrix K, as defined below:

[0080]

[0081] In the above formula, (u,v) are the coordinates of the pixel in the image coordinate system, (x... w ,y w ,z w Z represents the three-dimensional coordinates of a pixel in the camera coordinate system. c f is the depth from the 3D coordinates to the center of the image. x f y (X) is the focal length of the camera. c ,Y c Z c ) represents the position of the camera's center point in the world coordinate system, and α, β, and γ represent the phone's yaw, pitch, and roll angles, respectively.

[0082] In this embodiment, the coordinates of the key points in the close-up image in the world coordinate system are as follows:

[0083] Point S: [-0.27973470966848185, -0.2699330817189083, 1.466796875]

[0084] Point C: [-0.1595048574602088, -0.05785499396763191, 0.9013672471046448]

[0085] Point M: [-0.08165246985004983, -0.0009133816748723191, 0.463134765625];

[0086] The coordinates of key points in the distant image in the world coordinate system are as follows:

[0087] Female ear: [11.159053080437113, -9.793132763669858, 3.2738947643526237]

[0088] Aerial roots: [11.873333014660819, -9.936938614938233, 3.4883662015631045].

[0089] 9. Determine key points belonging to the same plant based on the Z and X value ranges of the key points' world coordinate system coordinates, and pair them up, wherein the Z and X value ranges are both less than 0.15 meters.

[0090] 10. Calculate the ear height and leaf angle of the maize plant based on the coordinates of the key points of pairing in the world coordinate system. The ear height is calculated using the following formula:

[0091]

[0092] In the above formula, EH represents the ear height. The three-dimensional coordinates of the key points of the maize female ear in the world coordinate system. The three-dimensional coordinates of the key points of corn aerial roots in the world coordinate system;

[0093] The leaf angle is calculated using the following formula:

[0094]

[0095] In the above formula, LA is the leaf angle, SM is the distance between points S and M, MC is the distance between points M and S, and SC is the distance between points S and C. These are the three-dimensional coordinates of points S, M, and C in the world coordinate system.

[0096] The ear height calculated in this embodiment is 0.759522182, while the measured value is 0.76, with a relative error of 0.06%; the leaf angle is 12.99°, while the measured value is 14.7°, with a relative error of 1.03%.

[0097] Example 2:

[0098] A field maize plant phenotypic measurement system, such as Figure 5 As shown, it includes an image acquisition unit, a YOLOv8 model, a key point pixel coordinate determination module, a key point camera coordinate determination module, a key point pairing module, and a module for calculating ear height and leaf angle.

[0099] The image acquisition device is a mobile phone equipped with a TOF sensor, which is used to capture RGB images and depth images of corn plants in the field. Both the RGB images and depth images include close-up and distant images.

[0100] The YOLOv8 model has a clip module added, which is located before the YOLOv8 backbone module;

[0101] The YOLOv8 model is used for:

[0102] The clip module determines the cut-off depth range based on the shooting distance and extracts pixels within the cut-off depth range from the depth image as a mask image. The cut-off depth range is [d]. min ,d min +0.3(d max -dmin )],d max d min These are the maximum and minimum depth values ​​that can be read from the depth image, respectively.

[0103] The extracted mask image is applied to the RGB image, and the RGB pixels covered by the mask image are extracted by image overlay analysis as the RGB image after mask processing.

[0104] Target recognition is performed on the RGB images after masking. For distant images, the target is identified as corn aerial roots and the segmented target is corn female ears. For close-up images, the target is identified as corn leaf nodes.

[0105] The key point pixel coordinate determination module is used to determine the key point pixel coordinates output by the identified target. Among them, the key point pixel coordinates of the corn aerial roots are the coordinates of the center point of the target box, the key point pixel coordinates of the corn female ear are the coordinates of the lowest point of the segmented region, and the key point pixel coordinates of the corn leaf node are the coordinates of point S on the leaf vein that is less than 50 pixels away from the leaf node, point M at the intersection of the leaf and the stem, and point C on the stem that is less than 50 pixels away from the leaf node.

[0106] The key point camera coordinate determination module is used for:

[0107] Extract the coordinate depth D corresponding to the pixel coordinates of the key point from the depth image, and determine the depth confidence of the point. If the depth confidence is lower than the threshold, replace the depth D with the average depth of multiple pixels around the point with confidence higher than the threshold. If the confidence of all pixels around the point is lower than the threshold, discard the depth image.

[0108] The camera coordinates of the key points were calculated.

[0109] The camera coordinate system coordinates of the key points and the attitude data of the image acquisition device are used to convert their coordinates into the world coordinate system. The key points belonging to the same plant are determined and paired by matching the Z and X values ​​of the world coordinate system coordinates. The range of the Z and X values ​​is less than 0.15 meters.

[0110] The ear height and leaf angle calculation module is used to calculate the ear height and leaf angle of the maize plant based on the world coordinate system coordinates of the key points of pairing.

[0111] The ear height is calculated using the following formula:

[0112]

[0113] In the above formula, EH represents the ear height. The three-dimensional coordinates of the key points of the female ear of maize. Three-dimensional coordinates of key points for corn aerial roots;

[0114] The blade angle is calculated using the following formula:

[0115]

[0116] In the above formula, LA is the leaf angle, SM is the distance between points S and M, MC is the distance between points M and S, and SC is the distance between points S and C. These are the three-dimensional coordinates of points S, M, and C, respectively.

Claims

1. A method for measuring the phenotype of a field corn plant, characterized in that, the method comprises: S1, using an image collector to take an RGB image and a depth image of the field corn plant, the RGB image and the depth image both containing close-range images and long-range images; S2, extracting a mask image from the depth image based on a clip module added in the YOLOv8 model, and applying the extracted mask image to the RGB image to extract the RGB pixels covered by the mask image as the RGB image after mask processing; S3, using the YOLOv8 model to perform target recognition on the RGB image after mask processing, wherein for the long-range image, the recognition target is corn aerenchyma, and the segmentation target is corn ear; for the close-range image, the recognition target is the corn node part; S4, determining the key point pixel coordinates output by the recognized target, wherein the key point pixel coordinates of the corn aerenchyma are the coordinates of the center point of the target frame, the key point pixel coordinates of the corn ear are the coordinates of the lowermost point of the segmentation region, and the key point pixel coordinates of the corn node part are the coordinates of a point S on the leaf vein which is less than 50 pixels away from the node, a point M at the intersection of the leaf and the stem, and a point C on the stem which is less than 50 pixels away from the node; S5, extracting the coordinate depth D corresponding to the key point pixel coordinates from the depth image, and calculating the camera coordinate system coordinates of the key point; S6, converting the camera coordinate system coordinates of the key point to world coordinate system coordinates according to the pose data of the image collector, determining the key points belonging to the same plant through coordinate matching, and pairing them; S7, calculating the ear height and leaf angle of the corn plant according to the world coordinate system coordinates of the paired key points.

2. The method for measuring the phenotype of a field corn plant according to claim 1, characterized in that, the clip module is located before the backbone module of YOLOv8; in S2, the extraction of the mask image from the depth image based on the clip module added in the YOLOv8 model comprises: The clip module first determines a clipping depth range according to the shooting distance, and then extracts pixels in the clipping depth range from the depth image as the mask image, where the clipping depth range is [d min , d min + 0.3(d max -d min )], d max , d min are respectively the maximum and minimum depth values that can be read in the depth image.

3. The method for measuring the phenotype of a field corn plant according to claim 1 or 2, characterized in that, in S5, while extracting the coordinate depth D corresponding to the key point pixel coordinates from the depth image, the depth confidence of the point is determined, if the depth confidence is lower than the threshold, the depth D is replaced by the average depth of the multiple pixel points around the point whose confidence is higher than the threshold, if the confidence of the pixel points around the point is all lower than the threshold, the depth image is discarded.

4. The method for measuring the phenotype of a field corn plant according to claim 1 or 2, characterized in that, in S7, the ear height is calculated according to the following formula: In the above formula, EH is the ear position height, is the three-dimensional coordinate of the key point of the corn ear, is the three-dimensional coordinate of the key point of the corn aerial root; the leaf angle is calculated according to the following formula: In the above formula, LA is the leaf included angle, SM is the distance between S point and M point, MC is the distance between M point and S point, SC is the distance between S point and C point, are three-dimensional coordinates of S point, M point and C point, respectively.

5. The method for measuring the phenotype of a field corn plant according to claim 1 or 2, characterized in that, The image collector is a mobile phone equipped with a TOF sensor, the distance between the image collector and the corn plant is not less than 0.75 m when shooting a long-range image, the distance between the image collector and the corn plant is 0.2 m when shooting a close-range image, and the pitch angle and roll angle of the image collector are both within the range of (-30°, 30°).

6. The field corn plant phenotype measurement method according to claim 1 or 2, wherein, The S6 determines the key points belonging to the same plant according to the Z value and X value ranges of the world coordinate system coordinates of the key points, and the Z value and X value ranges are both less than 0.15 m.

7. A field corn plant phenotype measurement system, comprising: The system comprises an image collector, a YOLOv8 model, a key point pixel coordinate determination module, a key point camera coordinate determination module, a key point pairing module, a ear height and leaf angle calculation module; The image collector is used to shoot RGB images and depth images of field corn plants, and the RGB images and depth images both contain close-range images and long-range images; The YOLOv8 model is used to: extract a mask image from the depth image based on the clip module added therein, and apply the extracted mask image to the RGB image to extract the RGB pixels covered by the mask image as the mask-processed RGB image; perform target recognition on the mask-processed RGB image, wherein for a long-range image, the recognition target is a corn aerenchyma, and the segmentation target is a corn ear; for a close-range image, the recognition target is a corn node part; The key point pixel coordinate determination module is used to determine the key point pixel coordinates output by the recognized target, wherein the key point pixel coordinates of the corn aerenchyma are the coordinates of the center point of the target frame, the key point pixel coordinates of the corn ear are the coordinates of the lowermost point of the segmentation region, and the key point pixel coordinates of the corn node part are the coordinates of a point S on the leaf vein which is less than 50 pixels away from the node, a point M at the intersection of the leaf and the stem, and a point C on the stem which is less than 50 pixels away from the node; The key point camera coordinate determination module is used to extract the coordinate depth D corresponding to the key point pixel coordinates from the depth image, and calculate the camera coordinate system coordinates of the key points in combination with the attitude and latitude and longitude data of the image collector; The key point pairing module is used to convert the camera coordinate system coordinates of the key points and the attitude data of the image collector into world coordinate system coordinates, determine the key points belonging to the same plant through coordinate matching, and pair the key points; The ear height and leaf angle calculation module is used to calculate the ear height and leaf angle of the corn plant according to the world coordinate system coordinates of the paired key points.

8. The field corn plant phenotype measurement system according to claim 7, wherein: The clip module is located before the backbone module of YOLOv8; The extraction of the mask image from the depth image comprises: The clip module first determines a clipping depth range according to the shooting distance, and then extracts pixels in the clipping depth range from the depth image as the mask image, where the clipping depth range is [d min , d min + 0.3(d max -d min )], d max , d min are respectively the maximum and minimum depth values that can be read in the depth image. The key point camera coordinate determination module extracts the coordinate depth D corresponding to the key point pixel coordinate from the depth image, and determines the depth confidence of the point at the same time, if the depth confidence is lower than the threshold, the depth D is replaced by the depth mean of the multiple pixel points around the point whose confidence is higher than the threshold, if the confidence of the pixel points around the point is all lower than the threshold, the depth image is discarded.

9. The field corn plant phenotype measurement system according to claim 7 or 8, wherein, The ear position height is calculated according to the following formula: In the above formula, EH is the ear position height, is the three-dimensional coordinate of the key point of the corn ear, is the three-dimensional coordinate of the key point of the corn aerial root. The leaf included angle is calculated according to the following formula: In the above formula, LA is the leaf included angle, SM is the distance between S point and M point, MC is the distance between M point and S point, SC is the distance between S point and C point, are three-dimensional coordinates of S point, M point and C point, respectively.

10. The field corn plant phenotype measurement system according to claim 7 or 8, wherein, The image collector is a mobile phone with a TOF sensor, when shooting a long shot image, the distance between the image collector and the corn plant is not less than 0.75 meters, when shooting a close shot image, the distance between the image collector and the corn plant is 0.2 meters, and the pitch angle and roll angle of the image collector are both within the range of (-30°, 30°).

Images