Platform and method for obtaining wheat tiller number during the entire growth period

Abstract

The present application discloses a platform and method for obtaining the number of tillers of wheat throughout its entire growth period, comprising: a self-propelled navigation vehicle, a cross slide mounted on the self-propelled navigation vehicle, and a combing mechanism mounted on the cross slide, wherein the combing mechanism is provided with a camera. When the number of tillers of wheat needs to be obtained, the combing mechanism combs out the wheat leaves and controls the camera to take pictures of the tiller position of the wheat root part, and the images obtained by taking pictures are used to determine the number of tillers. When the self-propelled navigation vehicle reaches a designated position to obtain the number of wheat tillers during the shaping of the wheat field, it controls the combing mechanism to comb out the wheat leaves so that the wheat leaves do not block the camera's shooting. After the camera has taken the image, the image is processed to obtain the number of wheat tillers. Due to the addition of the combing mechanism, the leaves that block the tiller angle of the wheat root part can be combed out, so that the captured image is not blocked by the leaves, and the acquisition of the number of wheat tillers in the later processing of the image is more accurate and faster.

Description

Technical Field

[0001] The present application relates to the technical field of agricultural equipment automation, and in particular to a platform and method for obtaining the number of tillers of wheat throughout its entire growth period. Background Art

[0002] Plant phenomics is the study of plant growth morphology, growth environment, and physiological and biochemical characteristics. Plant phenomics focuses on analyzing the role of various traits in a plant's life history. For wheat, tiller number is typically measured throughout the growing season. This data is valuable for studying wheat phenotypic data and holds important reference value for predicting wheat yield.

[0003] However, dense wheat plantings in the field, with overlapping leaves, make tiller number difficult to determine. Currently, there are few phenotyping devices specifically designed to measure wheat tiller number. These are primarily based on traditional plant physiology and genetic engineering methods, which measure parameters during plant growth and development to obtain relevant information. Traditional field methods for measuring wheat tiller number are time-consuming, labor-intensive, and offer low accuracy, often with some deviation from actual results.

[0004] Therefore, how to quickly and accurately obtain the number of wheat tillers is a technical problem that needs to be solved urgently in this field. Summary of the Invention

[0005] In order to solve the above technical problems, this application proposes the following technical solutions:

[0006] In the first aspect, an embodiment of the present application provides a platform for obtaining the number of tillers of wheat throughout the entire growth period, comprising: a self-propelled navigation vehicle, a cross slide installed on the self-propelled navigation vehicle, and a combing mechanism installed on the cross slide, wherein the combing mechanism is provided with a camera. When the number of tillers of wheat needs to be obtained, the combing mechanism combs open the wheat leaves and controls the camera to take pictures of the tiller position of the wheat root part, and the image obtained by the picture is used to determine the number of tillers.

[0007] In one possible implementation, the cross slide includes a transverse slide fixedly mounted on the self-propelled guided vehicle, the transverse slide being provided with a first slider and a first motor for controlling the movement of the first slider on the transverse slide, the transverse slide being slidably connected to the first slider; the first slider being provided with a second motor and a second slider, the second slider being slidably connected to the vertical slide, the second motor controlling the up and down movement of the vertical slide, and the lower end of the vertical slide being fixedly connected to the combing mechanism.

[0008] In one possible implementation, the combing mechanism includes a support arm, one end of which is provided with an end connecting plate, and the end connecting plate is fixedly connected to the vertical slide rail; the other end of the support arm is provided with a movable combing claw, and the movable combing claw is arranged around the end surface of the support arm, and a camera hole is provided in the middle of the movable combing claw.

[0009] In one possible implementation, the comb claw is supported by an outer ring connecting arm and an inner ring connecting arm, the inner ring connecting arm is arranged on the support arm through an inner ring fixing plate, the fixed end of the comb claw is rotatably connected to the inner ring connecting arm, the outer ring connecting arm is correspondingly arranged around the outside of the inner ring connecting arm, and is fixedly connected to the support arm through the outer ring fixing plate, and the outer ring connecting arm is rotatably connected to the comb claw.

[0010] In one possible implementation, a camera platform is provided at the other end of the support arm, and the inner ring connecting arm is fixedly provided on the camera platform. When the electric push rod pushes the camera platform, the camera platform pushes the inner ring connecting arm, and the comb claw opens with the cooperation of the outer ring connecting arm.

[0011] In one possible implementation, the self-propelled navigation vehicle is also provided with a controller, a display and a processor. The wheels of the self-propelled navigation vehicle are provided with motors, and the wheels are fixedly connected to the frame through the legs. The controller is used to control the walking of the self-propelled navigation vehicle and the operation of the combing mechanism, and the processor is used to process the captured images.

[0012] In a second aspect, an embodiment of the present application provides a method for obtaining the number of tillers of wheat throughout its entire growth period, based on the platform for obtaining the number of tillers of wheat throughout its entire growth period described in any possible implementation of the first aspect, the method comprising:

[0013] The controller controls the combing mechanism to be above the wheat, and the slide rail descends. The combing claws of the combing mechanism insert into the center of the wheat. The controller sends a signal and the end opens.

[0014] After the combing claws of the combing mechanism extend into the center of the wheat, the controller controls the opening and closing of the combing claws so that the combing claws are just opened to a degree that does not block the camera shooting; the combing claws comb the wheat leaves apart, so that the camera is facing the tiller position of the wheat root part and shoots the tillering part of the wheat;

[0015] After acquiring the image, the image is preprocessed by flipping it horizontally and vertically, enhancing it, and then binarizing it.

[0016] The skeleton extraction algorithm is used to perform edge detection and erosion on the binary image to extract the wheat tiller skeleton.

[0017] Perform morphological topological analysis on wheat skeleton images and quickly detect the intersection of skeleton curves through convex hull detection and convex defect detection;

[0018] By calculating and marking the defect points that meet the characteristics of wheat tillering angle, wheat tillering can be obtained.

[0019] In one possible implementation, the defect points that meet the characteristics of wheat tillering angle are marked by calculation, thereby achieving the acquisition of wheat tillers, including: tillering angle identification uses a refinement algorithm, first defining an action function for deleting pixels, re-modifying the values ​​of the iteratively selected pixel points from the initial setting to 0, repeatedly setting the deletion action according to the iterative idea, and adding a while loop to implement loop iteration; in the while loop, judging whether the deletion condition is met based on the sum of the number of neighboring pixels, and so on; when the neighborhood of the boundary pixel is less than three points, exiting the while loop and returning the refined pixel parameter value; calling the Skeleton function to apply the returned pixel parameters to the binary image, and displaying and saving it through the matplotlib function.

[0020] In one possible implementation, the HSV model is used to simulate the way colors appear under light, so as to adapt to the description of a single color under different light sources when identifying wheat tillers. The HSV spatial threshold segmentation is used to preprocess the image set, converting the original image into a binary image before performing tiller angle recognition and counting.

[0021] In one possible implementation, a three-dimensional model is used to describe the target image, and the values ​​of the three RGB color components are used as the coordinate values ​​of the ordinary Cartesian coordinate system in the Euclidean space. The X-axis represents red, which increases to the left, the Y-axis represents blue, which increases to the lower right, and the vertical z-axis represents green, which increases upward. After obtaining the image under the HSV channel, an area selection function is added to uniformly select the HSV channel data of the three points in the identification area according to the linear transformation, and output the corresponding matrix value.

[0022] In an embodiment of the present application, when a self-propelled navigation vehicle reaches a designated location during wheat field mapping to obtain the number of wheat tillers, it controls a combing mechanism to separate the wheat leaves so that they do not block the camera's image. After the camera captures the image, the image is processed to obtain the number of wheat tillers. The addition of the combing mechanism allows leaves that block the tiller angles of the wheat roots to be separated, ensuring that the captured image is not blocked by the leaves. This allows for more accurate and rapid acquisition of the wheat tiller count through subsequent image processing. BRIEF DESCRIPTION OF THE DRAWINGS

[0023] Figure 1 A schematic diagram of the structure of a platform for obtaining the number of tillers of wheat throughout its growth period provided in an embodiment of the present application;

[0024] Figure 2 A schematic diagram of the structure of the cross slide provided in an embodiment of the present application;

[0025] Figure 3 A closed schematic diagram of the combing mechanism provided in an embodiment of the present application;

[0026] Figure 4 A schematic diagram of an open combing mechanism provided in an embodiment of the present application;

[0027] Figure 5 A schematic flow chart of a method for obtaining the number of tillers during the entire growth period of wheat provided in an embodiment of the present application;

[0028] Figure 1-5 In the symbol, it is represented as:

[0029] 1-Self-propelled navigation vehicle, 2-Cross slide, 3-Combing mechanism, 4-Horizontal slide, 5-First slider, 6-First motor, 7-Second motor, 8-Second slider, 9-Vertical slide, 10-Support arm, 11-End connecting plate, 12-Movable combing claw, 13-Camera hole, 14-Outer ring connecting arm, 15-Inner ring connecting arm, 16-Inner ring fixing plate, 17-Outer ring fixing plate, 18-Camera platform, 19-Controller, 20-Display, 21-Processor, 22-Wheels, 23-Motor, 24-Legs. DETAILED DESCRIPTION

[0030] The present invention will be described below with reference to the accompanying drawings and specific implementation methods.

[0031] Figure 1 This is a schematic diagram of a wheat tiller number acquisition platform for the entire growth period provided in the embodiment of the present application, see Figure 1 The platform for obtaining the number of tillers of wheat during the entire growth period provided in this embodiment includes: a self-propelled navigation vehicle 1, a cross slide 2 installed on the self-propelled navigation vehicle 1, and a combing mechanism 3 installed on the cross slide 2. The combing mechanism 3 is provided with a camera. When the number of tillers of wheat needs to be obtained, the combing mechanism 3 combs open the wheat leaves and controls the camera to take pictures of the tiller position of the wheat root part. The image obtained by the photo is used to determine the number of tillers.

[0032] See also Figure 2The cross slide 2 includes a transverse slide 4 fixedly mounted on the self-propelled guided vehicle 1. The transverse slide 4 is provided with a first slider 5 and a first motor 236 for controlling the movement of the first slider 5 on the transverse slide 4. The transverse slide 4 is slidably connected to the first slider 5. The first slider 5 is provided with a second motor 237 and a second slider 8. The second slider 8 is slidably connected to the vertical slide 9. The second motor 237 controls the up and down movement of the vertical slide 9. The lower end of the vertical slide 9 is fixedly connected to the combing mechanism 3. The transverse slide 4 is linked to the track through the base, and the transverse movement of the first slider 5 on the track is achieved by the first motor 236. The second motor 237 drives the vertical slide 9 to move longitudinally. The base is used to connect different sensors to achieve image acquisition requirements of different sensors.

[0033] Specifically, the sliding mechanism consists of a track, a slider, a synchronous wheel shaft, and a reduction gear. A gap exists between the track and the slider. The inner side of the track and the outer side of the slider are connected to the fixed member via a connecting rod parallel to the inner side of the slider. The bottom of the slider is connected to an adjacent fixed member. When the sliding mechanism exchanges data with other devices, it will not deviate or change its direction of movement.

[0034] See also Figure 3 and Figure 4 The combing mechanism 3 includes a support arm 10, one end of which is provided with an end connecting plate 11, and the end connecting plate 11 is fixedly connected to the vertical slide rail 9; the other end of the support arm 10 is provided with a movable combing claw 12, and the movable combing claw 12 is arranged around the end surface of the support arm 10, and a camera hole 13 is provided in the middle of the movable combing claw 12.

[0035] The comb claw is supported by an outer ring connecting arm 14 and an inner ring connecting arm 15. The inner ring connecting arm 15 is mounted on the support arm 10 via an inner ring fixing plate 16. The fixed end of the comb claw is rotatably connected to the inner ring connecting arm 15. The outer ring connecting arm 14 is correspondingly arranged around the outside of the inner ring connecting arm 15 and is fixedly connected to the support arm 10 via an outer ring fixing plate 17. The outer ring connecting arm 14 is rotatably connected to the comb claw. A camera platform 18 is provided at the other end of the support arm 10. The inner ring connecting arm 15 is fixedly mounted on the camera platform 18. When the electric push rod pushes the camera platform 18, the camera platform 18 pushes the inner ring connecting arm 15, and the comb claw opens with the cooperation of the outer ring connecting arm 14.

[0036] An electric push rod inside the support arm 10 pushes the camera platform 18 to open and close the movable combing claws 12. Driven by the push rod and the connecting arm, the camera platform 18 translates horizontally. When the movable combing claws 12 are fully open, the camera platform 18 reaches the optimal image capture position, avoiding obstruction by the movable combing claws 12. The comb teeth formed on the movable combing claws 12 act as a combing mechanism, separating the wheat leaves for better image capture.

[0037] The self-propelled navigation vehicle 1 is also provided with a controller 19, a display 20 and a processor 21. The wheels 22 of the self-propelled navigation vehicle 1 are provided with motors 23, and the wheels 22 are fixedly connected to the frame through the legs 24. The controller 19 is used to control the walking of the self-propelled navigation vehicle 1 and the operation of the combing mechanism 3, and the processor 21 is used to process the captured images.

[0038] The wheels of the self-propelled guided vehicle (1) are specially designed for field environments, reducing ground contact area for easier steering and preventing damage to wheat. The wheels (22) are also spaced farther apart to prevent slipping. The vehicle's overall height can be adjusted for different growth stages, facilitating phenotypic data collection. The front-end clamping device accommodates sensor capture requiring higher altitudes, such as hyperspectral cameras.

[0039] The motors 23 mounted on the legs 24 enable the platform to maintain four-wheel drive throughout the entire driving process. The engine output torque is distributed to the front and rear wheels in a fixed ratio. This driving mode provides excellent off-road and maneuverability at all times. Combined with the specially designed field wheels, it is more suitable for traveling between ridges. The wheels can rotate 360 ​​degrees to achieve horizontal movement on the edge of the field without turning, improving efficiency and reducing damage to the wheat.

[0040] The four-wheel drive system utilizes the entire vehicle's weight as adhesion pressure, significantly increasing adhesion and extending traction limits. The power from the motor 23 is transmitted individually to each wheel, reducing the driving force burden on each drive wheel. This ensures sufficient power is delivered to the road without exceeding the tire's friction limit (preventing wheel slip). This results in uniform tire wear, extending tire life.

[0041] The front self-balancing device uses a built-in precision solid-state gyroscope to determine the vehicle's posture. A high-speed, precise central microprocessor 21 calculates appropriate instructions and drives the motor to achieve balance. This is primarily based on a fundamental principle known as "dynamic stability," which is the vehicle's inherent self-balancing capability.

[0042] The intelligent control device controls the overall route of the autonomous navigation vehicle. After completing the path planning based on the field conditions, the vehicle's sensors identify the field coordinates and adjust the route. Upon reaching the designated location, a command is issued to the cross rail 2, adjusting the vertical rails to capture wheat image information. After the image is captured, it is sent to the processor 21 and fed back to the intelligent control device. The rails are retracted, and a travel command is issued to the wheel motors 23, driving the four-wheel drive autonomous vehicle forward to complete the next set of operations.

[0043] Corresponding to the platform for obtaining the number of tillers of wheat throughout its entire growth period provided in the above embodiment, the present application also provides an embodiment of a method for obtaining the number of tillers of wheat throughout its entire growth period.

[0044] See also Figure 5 The method for obtaining the number of tillers during the entire growth period of wheat in this embodiment includes:

[0045] S101, the controller controls the slide rail to descend when the combing mechanism is located above the wheat, the combing claws of the combing mechanism are inserted into the center of the wheat, the controller sends a signal, and the end opens.

[0046] S102, after the combing claws of the combing mechanism extend into the center of the wheat, the controller controls the opening and closing degree of the combing claws so that the combing claws are opened to a degree that does not block the camera shooting; the combing claws comb the wheat leaves apart, so that the camera is facing the tiller position of the wheat root part, and the tillering part of the wheat is photographed.

[0047] S103, after acquiring the image, pre-processing the image is first performed, such as flipping the image horizontally and vertically, enhancing the image, and then binarizing the image.

[0048] S104, performing edge detection and corrosion on the binary image using a skeleton extraction algorithm to extract the wheat tiller skeleton.

[0049] S105 , performing morphological topological analysis on the wheat skeleton image, and quickly detecting the intersection of the skeleton curves through convex hull detection and convex defect detection.

[0050] S106, marking defect points that meet the characteristics of wheat tillering angles by calculation, thereby achieving the acquisition of wheat tillers.

[0051] Based on the principles of the thinning algorithm, a thinning program for wheat binary images was designed. To determine the relationship between foreground and background pixels and better calibrate the image boundary, after reading the binary image using the io.imread function, the neighborhood pixels were defined. For each iteration of the boundary pixels, the pixel values ​​were returned in clockwise order. After defining the pixel boundaries, the main program was designed. According to the thinning algorithm, the action function for deleting pixels was first defined. The values ​​of the pixels selected through iteration were reset from the initial setting to 0. Based on the iterative concept, the deletion action was repeatedly set, and a while loop was added to implement loop iteration. In the while loop, the sum of the number of neighborhood pixels was used to determine whether the deletion condition was met, and so on. When the neighborhood of a boundary pixel was less than three points, the while loop was exited and the refined pixel parameter value was returned. The Skeleton function was called to apply the returned pixel parameters to the binary image, and the result was displayed and saved using the matplotlib function.

[0052] While skeleton detection of wheat during the tillering period, designed based on a refinement algorithm, has proven efficient and accurate in validating most model results, the independently designed refinement algorithm parameters cannot compensate for the lack of a theoretical foundation. Each design for practical application requires extensive experience and precise parameter control. Therefore, a linear spanning network (LSU) based on a linear spanning unit (LSU) structure was introduced to perform edge detection and skeletonization on binary images of wheat during the tillering period. Skeleton images contain concise and clear visual information. As the most representative aspect of morphology, they have significantly advanced the field of machine vision. Deep learning is achieving unprecedented results in various fields. Similarly, in machine vision, binary image skeletonization using convolutional neural networks (CNNs) as the primary layer has also demonstrated remarkable success.

[0053] Linear span networks (LSNs) are commonly used for object skeletonization detection. This method, based on the structure of a fully nested edge detection (HED) network, incorporates a linear span unit (LSU). This method integrates multiple data features, fusing linear span features with resolution matching and subspace linear spans. This results in improved noise immunity in the output binary image, enabling filtering of background noise that is closer to reality. This provides superior edge feature extraction performance and also demonstrates good performance in generating morphological skeletons. With the help of convolutional neural network layers (VGG-16), the skeletonization of the LSN is more adaptable to morphological objects of varying dimensions.

[0054] After generating a skeleton image of wheat during the tillering phase using a linear spanning network, a morphological topological analysis of the wheat skeleton image is performed. Based on the characteristics of the skeleton image, this paper proposes a method that combines convex hull detection and convex defect detection to perform graphical analysis of the wheat tiller skeleton image. This method can quickly detect the intersection of skeleton curves and, through calculation, mark defect points that match the wheat tiller angle characteristics, thereby achieving accurate wheat tiller angle counting.

[0055] In this embodiment, the HSV model is used to simulate the display mode of color under light to adapt to the description of a single color under different light sources when identifying wheat tillers; the HSV spatial threshold segmentation is used to preprocess the image set, and the original image is converted into a binary image before the tiller angle identification and counting work is performed.

[0056] In traditional image processing, the RGB model is often used to analyze and process images. However, when it comes to wheat tillering recognition, the RGB model is unable to cope with the situation where wheat is a single color due to its low level of description of single colors. Furthermore, the RGB model lacks descriptions of color saturation and brightness, making it unsuitable for realistic situations with varying light sources. Therefore, the RGB model was abandoned.

[0057] The HSV model used is a representation of midpoints in the RGB color model in a cylindrical coordinate system, simulating how colors appear under light. In the HSV model, H represents hue, S represents saturation, and V represents value. The HSV color space separates value (V) and saturation (S) from color data, minimizing the correlation between individual colors. This makes it suitable for describing a single color under different light sources when identifying wheat tillers.

[0058] A three-dimensional model is used to describe the target image. The values ​​of the three RGB color components are used as the coordinate values ​​of the ordinary Cartesian coordinate system in the Euclidean space. The X-axis represents red, increasing to the left, the Y-axis represents blue, increasing to the lower right, and the vertical z-axis represents green, increasing upward. After obtaining the image under the HSV channel, an area selection function is added to uniformly select the HSV channel data of the three points in the recognition area according to the linear transformation, and output the corresponding matrix value.

[0059] In the embodiments of the present application, "at least one" refers to one or more, and "more" refers to two or more. "And / or" describes the association relationship of associated objects, indicating that three relationships may exist. For example, A and / or B can represent the existence of A alone, the existence of A and B at the same time, and the existence of B alone. Among them, A and B can be singular or plural. The character " / " generally indicates that the previous and next associated objects are in an "or" relationship. "At least one of the following" and similar expressions refer to any combination of these items, including any combination of single or plural items. For example, at least one of a, b and c can be represented by: a, b, c, ab, ac, bc, or abc, where a, b, c can be single or multiple.

[0060] The above description is merely a specific embodiment of the present application. Any person skilled in the art may easily conceive of variations or substitutions within the technical scope disclosed in this application, and such variations or substitutions shall be within the scope of protection of this application. The scope of protection of this application shall be subject to the scope of protection of the claims.

Claims

1. A platform for obtaining the number of tillers during the entire growth period of wheat, characterized in that: include: A self-propelled navigation vehicle, a cross slide mounted on the self-propelled navigation vehicle, and a combing mechanism mounted on the cross slide, wherein the combing mechanism is provided with a camera. When the number of tillers of wheat needs to be obtained, the combing mechanism combs the wheat leaves apart and controls the camera to take a picture of the tiller position of the wheat root part. The image obtained by the picture is used to determine the number of tillers. The cross slide includes a transverse slide fixedly arranged on the self-propelled guided vehicle, the transverse slide is provided with a first slider and a first motor that controls the movement of the first slider on the transverse slide, the transverse slide is slidably connected to the first slider; the first slider is provided with a second motor and a second slider, the second slider is slidably connected to the vertical slide, the second motor controls the vertical slide to move up and down, and the lower end of the vertical slide is fixedly connected to the combing mechanism; The combing mechanism includes a support arm, one end of which is provided with an end connecting plate, and the end connecting plate is fixedly connected to the vertical slide rail; the other end of the support arm is provided with a movable combing claw, and the movable combing claw is arranged around the end surface of the support arm, and a camera hole is provided in the middle of the movable combing claw.

2. The wheat tiller number acquisition platform for the entire growth period according to claim 1, characterized in that: The movable comb claw is supported by an outer ring connecting arm and an inner ring connecting arm. The inner ring connecting arm is arranged on the support arm through an inner ring fixing plate. The fixed end of the movable comb claw is rotatably connected to the inner ring connecting arm. The outer ring connecting arm is correspondingly arranged around the outside of the inner ring connecting arm and is fixedly connected to the support arm through the outer ring fixing plate. The outer ring connecting arm is rotatably connected to the movable comb claw.

3. The wheat tiller number acquisition platform for the entire growth period according to claim 2, characterized in that: A camera platform is provided at the other end of the support arm, and the inner ring connecting arm is fixedly provided on the camera platform. When the electric push rod pushes the camera platform, the camera platform pushes the inner ring connecting arm, and the movable comb claw opens with the cooperation of the outer ring connecting arm.

4. The platform for obtaining the number of tillers of wheat during the entire growth period according to any one of claims 1 to 3, characterized in that: The self-propelled navigation vehicle is also provided with a controller, a display and a processor. The wheels of the self-propelled navigation vehicle are provided with motors, and the wheels are fixedly connected to the frame through the legs. The controller is used to control the walking of the self-propelled navigation vehicle and the operation of the combing mechanism, and the processor is used to process the captured images.

5. A method for obtaining the number of tillers in the entire growth period of wheat, characterized in that: Based on the wheat tiller number acquisition platform for the entire growth period according to any one of claims 1 to 4, the method comprises: The controller controls the combing mechanism to be above the wheat, and the slide rail descends. The movable comb claws of the combing mechanism insert into the center of the wheat. The controller sends a signal and the end opens. After the combing mechanism's movable comb claws extend into the center of the wheat, the controller controls the opening and closing of the movable comb claws so that the movable comb claws are opened to a degree that does not block the camera's shooting. The movable comb claws comb the wheat leaves apart, so that the camera is facing the tiller position of the wheat root part and shoots the tillering part of the wheat. After acquiring the image, the image is preprocessed, the image is flipped horizontally and vertically, the image is enhanced, and then the image is binarized; The skeleton extraction algorithm is used to perform edge detection and erosion on the binary image to extract the wheat tiller skeleton. Perform morphological topological analysis on wheat skeleton images and quickly detect the intersection of skeleton curves through convex hull detection and convex defect detection; By calculating and marking the defect points that meet the characteristics of wheat tillering angle, wheat tillering can be obtained.

6. The method for obtaining the number of tillers in the whole growth period of wheat according to claim 5, wherein: The method marks defect points that meet the characteristics of wheat tillering angles by calculation, thereby achieving the acquisition of wheat tillers, including: using a refinement algorithm for tillering angle identification, first defining an action function for deleting pixels, resetting the values ​​of iteratively selected pixel points from an initial setting to 0, repeatedly setting the deletion action based on the iterative idea, and adding a while loop to achieve loop iteration; in the while loop, judging whether the deletion condition is met based on the sum of the number of neighboring pixels, and so on; when the neighborhood of a boundary pixel is less than three points, exiting the while loop and returning the refined pixel parameter value; calling the Skeleton function to apply the returned pixel parameter to a binary image, and displaying and saving it through the matplotlib function.

7. The method for obtaining the number of tillers in the whole growth period of wheat according to claim 6, wherein: The HSV model is used to simulate the display of color under light to adapt to the description of a single color under different light sources when identifying wheat tillers. The HSV spatial threshold segmentation is used to preprocess the image set, converting the original image into a binary image before performing tiller angle recognition and counting.

8. The method for obtaining the number of tillers in the whole growth period of wheat according to claim 7, wherein: A three-dimensional model is used to describe the target image. The values ​​of the three RGB color components are used as the coordinate values ​​of the ordinary Cartesian coordinate system in the Euclidean space. The X-axis represents red, increasing to the left, the Y-axis represents blue, increasing to the lower right, and the vertical z-axis represents green, increasing upward. After obtaining the image under the HSV channel, an area selection function is added to uniformly select the HSV channel data of the three points in the recognition area according to the linear transformation, and output the corresponding matrix value.

Images