Accurate alignment secondary regulation weeding machine and alignment method

Abstract

The application discloses a precision inter-row and inter-plant secondary regulation weeding machine, wherein the main beam is fixed with the rear frame of a tractor, the front and rear ends of a primary inter-row mechanism are respectively fixed with the main beam and the main frame, the front and rear ends of a secondary row distance regulation mechanism are respectively fixed with the main frame and inter-plant weeding monomers, a primary camera and a secondary camera are respectively fixed at the left and right ends of the main frame, the primary camera is downwardly shot, and the secondary camera is forwardly shot; the number of the inter-plant weeding monomers is at least one group, and the number of the secondary row distance regulation mechanism is consistent with that of the inter-plant weeding monomers. The application simultaneously realizes inter-row and inter-plant weeding, and through the precision positioning technology and the intelligent control system, the weeding operation is more accurate, the misoperation to crops is effectively reduced, especially the precision inter-plant weed control, and the safety of the corn root system is protected.

Description

TECHNICAL FIELD

[0001] The application belongs to the technical field of weeding, and specifically relates to a precision row-by-row secondary regulation weeding machine. BACKGROUND

[0002] Weeding in agricultural production is an important link to ensure the healthy growth of crops. Traditional weeding methods mainly include manual weeding and chemical weeding, but these methods are either labor-intensive and inefficient or harmful to the environment. With the development of technology, precision agriculture technology has emerged, aiming to improve the efficiency and sustainability of agricultural production. Intelligent weeding technology, especially weeding robots based on image recognition and machine learning, provides a new solution to reduce the use of chemical herbicides, reduce labor demand, and improve weeding efficiency. However, how to ensure the safety of crops while achieving efficient and precise weeding between rows and between plants is still a major challenge for existing technology. Therefore, it is necessary to provide an intelligent weeding machine that can accurately and efficiently remove weeds between rows and between plants. SUMMARY

[0003] In view of the problems in the background art, the application provides a precision row-by-row secondary regulation weeding machine, and the technical scheme is as follows: comprising a main beam, a primary row-by-row mechanism, a main frame, a secondary row distance adjustment mechanism, and a row-plant weeding unit, wherein the main beam is fixed to the rear frame of a tractor, the front and rear ends of the primary row-by-row mechanism are fixed to the main beam and the main frame respectively, the front and rear ends of the secondary row distance adjustment mechanism are fixed to the main frame and the row-plant weeding unit respectively; a primary camera and a secondary camera are fixed to the left and right ends of the main frame respectively, the primary camera shoots forward, and the secondary camera shoots downward; the number of row-plant weeding units is at least one group, and the number of secondary row distance adjustment mechanisms is consistent with the number of row-plant weeding units.

[0004] The primary row-by-row mechanism comprises: two rhombic suspensions, a suspension front mounting bracket, a suspension rear mounting bracket, and a hydraulic cylinder, the front and rear ends of the two rhombic suspensions are hinged to the suspension front mounting bracket and the suspension rear mounting bracket respectively as the front and rear ends of the primary row-by-row mechanism, and the connecting line of the front and rear hinge points of the rhombic suspension is parallel to the connecting line of the front and rear hinge points of the other rhombic suspension; an angle sensor is arranged at any hinge point of the two rhombic suspensions; one end of the hydraulic cylinder is fixed to the main beam, and the other end of the hydraulic cylinder is connected to the rhombic suspension provided with the angle sensor, and the hydraulic cylinder is used to push the relative displacement between the main beam and the main frame.

[0005] The secondary row spacing adjustment mechanism comprises an electric push rod, a pulley mounting plate, a connecting rod and a weeding unit fixing plate, wherein the shell of the electric push rod is fixed as the front end of the secondary row spacing adjustment mechanism with the main frame, the movable end of the electric push rod is fixed with the weeding unit fixing plate through the connecting rod, two sections of the two pulley mounting plates are respectively fixed with a weeding unit fixing plate, the pulley mounting plate is provided with two groups of side shift pulleys through deep groove ball bearings and pulley mounting shafts, and the two groups of side shift pulleys are in contact with the main frame; the weeding unit fixing plate is fixed as the rear end of the secondary row spacing adjustment mechanism with the inter-row and inter-plant weeding unit.

[0006] The inter-row and inter-plant weeding unit comprises a weeding equipment frame, an inter-row weeding production combination, an inter-plant flexible weeding finger wheel, a profiling mechanism and a roller, wherein the two ends of the profiling mechanism are respectively fixed with the weeding unit fixing plate and the weeding equipment frame, the inter-row weeding production combination and the inter-plant flexible weeding finger wheel are fixedly installed below the weeding equipment frame, and the roller is installed on the weeding equipment frame and located in front of the inter-row weeding production combination.

[0007] The inter-row weeding production combination comprises two A-type soil dividing shovels for removing weeds in the inter-row of crops.

[0008] The main frame comprises two front and rear beams for respectively installing the primary alignment mechanism and the secondary row spacing adjustment mechanism.

[0009] An alignment method of the precision alignment secondary regulation weeding machine is also provided, and the technical scheme is as follows:

[0010] Step 1, reset, initially, the primary alignment mechanism and the secondary row spacing adjustment mechanism are in initial positions;

[0011] Step 2, initialization adjustment is performed on the primary alignment mechanism, including:

[0012] Step 21, the primary camera remotely photographs crop rows and extracts an error angle as a moving angle θ,

[0013] Step 22, the hydraulic cylinder in the primary alignment mechanism pushes the parallel cantilever to make the center position line aligned with the navigation line; the primary regulation is the lengthening / shrinking amount ΔD1 of the hydraulic cylinder, ΔD1 = the moving angle θ × the cantilever length L, and the cantilever length L is an input value;

[0014] Step 3, initialization adjustment is performed on the secondary row spacing adjustment mechanism, including:

[0015] Step 31, the secondary camera photographs crop rows from below and fits four crop row lines,

[0016] Step 32, taking the position of the first column of corn rows as zero point, obtaining the position coordinate information of the second column, the third column and the fourth column of crop rows respectively according to the row line information, and adjusting the intersection line of the flexible weeding finger in the second column, the third column and the fourth column to align with the navigation line; the secondary regulation is the extension / shrinkage amount ΔS2 of the electric push rod, ΔS2 = electric push rod speed v × time t;

[0017] Step 4, after determining the row spacing, the error angle is tracked by the primary camera during the working of the weeding machine to keep the center line aligned with the navigation line.

[0018] Before step 1, calibration is performed, taking the weeding center line as the position of the center of the weeding component, and then inputting the spacing between each inter-row weeding product combination as the spacing between the actual positions of each column of weeding components, and resetting all electric push rods.

[0019] The method for extracting the error angle in steps 21 and 4 includes:

[0020] Step 211. Input the image captured by the primary camera;

[0021] Step 212. Perform grayscale conversion, Gaussian blur and edge detection on the image captured by the primary camera in sequence;

[0022] Step 213. ROI acquisition and perspective change: acquire the region of interest (ROI) according to the width and height of the image and a preset ratio, and apply perspective transformation to the ROI, so as to change the image from the original view to the bird's eye view;

[0023] Step 214. Calculate the histogram and find the histogram peak value: calculate the histogram of the transformed image, and find the peak values on the left and right sides, which correspond to the positions of the left and right sides of the crop row;

[0024] Step 215. Find the crop row using the sliding window method: use the histogram peak value as the starting point, and apply the sliding window method upward to find the pixel index of the entire crop row, and through these pixels, a polynomial curve representing the crop row can be fitted;

[0025] Step 216. Polynomial curve fitting and crop row: use the found pixel points to fit a polynomial curve to estimate the path of the crop row;

[0026] Step 217. Inverse perspective transformation: inverse perspective transform the bird's eye view containing the estimated crop row back to the original view;

[0027] Step 218. Calculate the evaluation line center point: at the height of the bottom of the image, calculate the left and right boundary points of the crop row, and then calculate the center position of the two boundary points as the evaluation line center point, which represents the current position of the machine relative to the crop row;

[0028] Step 219. Calculate the target center point: at another higher position in the image, repeat the above process to calculate the left and right boundary points of the crop row and the center position of the two boundary points, which represents the expected position of the machine at a future time point, that is, if the robot or vehicle maintains the current heading, it will reach this position;

[0029] Step 2110 Calculate the error angle: fit the evaluation line center points and the target center points to form two straight lines: one is the current travel direction, and the other is the expected travel direction; the heading angle error is the included angle between the two straight lines.

[0030] The fitting method of the crop row line in step 31 includes:

[0031] Step 311, take a picture by a secondary camera;

[0032] Step 312, use Unet semantics to segment the image to extract the posture of each row of corn seedlings;

[0033] Step 313, convert the semantic segmentation result into a gray image;

[0034] Step 314, perform Gaussian blur processing on the gray image to reduce noise and then perform edge detection;

[0035] Step 315, detect the corner points in the blurred image;

[0036] Step 316, draw a vertical line as the crop row line according to the x-coordinate of the feature points.

[0037] The beneficial effects of the present application are:

[0038] 1. Simultaneous inter-row and inter-plant weed control: the weed control machine can simultaneously target weeds in the inter-row and inter-plant spaces, thereby effectively controlling weeds and greatly improving the comprehensiveness and efficiency of weed control; at the same time, the automatic weed control machine greatly reduces the labor intensity and cost of manual weed control, meeting the needs of large-scale agricultural production.

[0039] 2. Improved weed control precision and reduced crop damage: precise positioning technology and intelligent control systems enable more precise weed control operations, effectively reducing damage to crops, especially for inter-plant weed control, which protects the safety of corn root systems.

[0040] 3. Reduce the dependence on chemical herbicides: by physical method of weeding, reduce the use of chemical herbicides, environmentally friendly, reduce the chemical pollution to crops and soil, improve the safety of agricultural products.

[0041] 4. Improve crop yield: by effective weeding, reduce the competition of weeds for water, light, nutrition, help to improve the growth conditions and yield of crops. BRIEF DESCRIPTION OF DRAWINGS

[0042] Figure 1 The flowchart of the embodiment of the precision row-by-row secondary regulation weeding machine of the application is shown in the figure;

[0043] Figure 2 The structure diagram of the primary row-by-row mechanism in the embodiment of the application is shown in the figure;

[0044] Figure 3 The structure diagram of the single secondary row spacing adjustment mechanism in the embodiment of the application is shown in the figure;

[0045] Figure 4 The structure diagram of the single secondary row spacing adjustment mechanism in the embodiment of the application is shown in the figure; Figure 3 The cross-sectional view of A-A section is shown in the figure;

[0046] Figure 5 The structure diagram of the row-by-row weeding monomer in the embodiment of the application is shown in the figure;

[0047] Figure 6 The flowchart of the working process of the embodiment of the application is shown in the figure;

[0048] Figure 7 The working process diagram of the control system in the embodiment of the application is shown in the figure;

[0049] Figure 8 The flowchart of the initialization adjustment of the primary row-by-row mechanism in the embodiment of the application is shown in the figure;

[0050] Figure 9 The initial position and the theoretical position of the primary row-by-row mechanism in the embodiment of the application are shown in the figure;

[0051] Figure 10 The flowchart of the initialization adjustment of the secondary row spacing adjustment mechanism in the embodiment of the application is shown in the figure;

[0052] Figure 11 The position diagram of the fitted crop row line in the embodiment of the application is shown in the figure;

[0053] Figure 12 The initial position and the theoretical position of the secondary row-by-row mechanism in the embodiment of the application are shown in the figure.

[0054] Among them, 1-main frame, 2-first-level row alignment mechanism, 3-inter-row weeding unit, 4-second-level row spacing adjustment mechanism, 5-second-level camera, 6-first-level camera, 7-electric push rod, 8-pulley mounting plate, 9-connecting rod, 10-weeding unit fixing plate, 11-side-moving pulley, 12-deep groove ball bearing, 13-pulley mounting shaft, 14-diamond cantilever, 15-hydraulic cylinder, 16-inter-row flexible weeding finger wheel, 17-contour mechanism, 18-pressing wheel, 19-weeding equipment frame, 21-cantilever front mounting frame, 20-inter-row weeding unit assembly, 22-cantilever rear mounting frame. Detailed Implementation

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

[0056] like Figure 1 The embodiment of the present invention shown includes: a main beam 19, a primary row alignment mechanism 2, a main frame 1, a secondary row spacing adjustment mechanism 4, and inter-row weeding units 3. The main beam 19 is fixed to the rear frame of the tractor. The front and rear ends of the primary row alignment mechanism 2 are fixed to the main beam 19 and the main frame 1, respectively. The front and rear ends of the secondary row spacing adjustment mechanism 4 are fixed to the main frame 1 and the inter-row weeding units 3, respectively. The front and rear ends of the primary row alignment mechanism 2 and the secondary row spacing adjustment mechanism 4 can move horizontally relative to each other and are fixed in their relative positions. A primary camera 6 and a secondary camera 5 are fixed to the left and right ends of the main frame 1, respectively. The primary camera 6 shoots forward, and the secondary camera 5 shoots downward. The number of inter-row weeding units 3 is at least one set, and the number of secondary row spacing adjustment mechanisms 4 is the same as the number of inter-row weeding units 3; in this embodiment, there are four sets.

[0057] In this embodiment, the main frame 1 includes two beams, front and rear, which are used to install the primary alignment mechanism 2 and the secondary row spacing adjustment mechanism 4, respectively. The two beams are fixed together by a connecting rod.

[0058] like Figure 1 and Figure 2 The primary alignment mechanism shown mainly consists of a rhomboid cantilever 14, a front cantilever mounting bracket 21, a rear cantilever mounting bracket 22, and a hydraulic cylinder 15. The front and rear ends of the two rhomboid cantilever 14 serve as the front and rear ends of the primary alignment mechanism, respectively hinged to the front and rear cantilever mounting brackets. The line connecting the two front and rear hinge points of the rhomboid cantilever 14 is parallel to the line connecting the two front and rear hinge points of the other rhomboid cantilever 14. An angle sensor (used to match the movement angle θ of the primary mechanism) is installed at any hinge point of the two rhomboid cantilever 14. One end of the hydraulic cylinder is fixed to the main beam 19, and the other end of the hydraulic cylinder is connected to the rhomboid cantilever 14 on which the angle sensor is installed, so as to generate relative displacement between the main beam 19 and the main frame 1.

[0059] like Figure 1 , Figure 3 andFigure 4 The secondary row spacing adjustment mechanism 4 shown mainly consists of an electric push rod 7, a pulley mounting plate 8, a connecting rod 9 and a weeding unit fixing plate 10, wherein the shell of the electric push rod is fixed as the front end of the secondary row spacing adjustment mechanism 4 with the main frame 1, the movable end of the electric push rod 7 is fixed with the weeding unit fixing plate 10 through the connecting rod 9, two sections of the two pulley mounting plates 8 are respectively fixed with a weeding unit fixing plate 10, the pulley mounting plate 8 is installed with two groups of side shift pulleys 11 in parallel through the deep groove ball bearing 12 and the pulley mounting shaft 13, the two groups of side shift pulleys 11 are respectively in contact with one side of the main frame 1, so that the secondary row spacing adjustment mechanism 4 can hold the main frame 1 (the rear beam in this embodiment) and produce relative sliding; the weeding unit fixing plate 10 is fixed as the rear end of the secondary row spacing adjustment mechanism 4 with the inter-row and inter-plant weeding unit 3;

[0060] As shown in Figure 5 The inter-row and inter-plant weeding unit 3 shown mainly consists of a weeding equipment frame 18, an inter-row weeding production combination 20, an inter-plant flexible weeding finger wheel 15, a profiling mechanism 16 and a press wheel 17, wherein the two ends of the profiling mechanism 16 are respectively fixed with the weeding unit fixing plate 10 and the weeding equipment frame 18, the inter-row weeding production combination 20 and the inter-plant flexible weeding finger wheel 15 are fixedly installed below the weeding equipment frame 18, and the press wheel 17 is installed on the weeding equipment frame 18 and located in front of the inter-row weeding production combination 20;

[0061] The inter-row weeding production combination 20 is responsible for removing weeds in the inter-row of crops, and the inter-row weeding production combination 20 includes two A-type soil separating shovels; since the weed roots are not as stable as the corn roots, when the two inter-plant flexible weeding finger wheels 15 continuously rotate the weeding tines in the inter-plant, the weeds in the inter-plant will be separated from the soil, thereby achieving the purpose of removing weeds in the inter-plant;

[0062] In this embodiment, the profiling mechanism 16 with a main structure of a parallelogram frame structure maintains the inter-row weeding production combination 20 and the inter-plant flexible weeding finger wheel 15 at a certain depth of entry into the ground through the elastic force of the profiling spring therein, the two metal beams opposite to each other in front and back of the parallelogram frame structure are the two ends of the profiling mechanism 16, and the profiling spring is installed between any two metal beams of the parallelogram frame structure.

[0063] As shown in Figure 6 and Figure 7 The alignment method used when the embodiment works is as follows:

[0064] Step 1, reset, initially the primary row aligning mechanism and the secondary row spacing adjustment mechanism are in the initial position;

[0065] Step 2, initialize adjustment is performed on the primary row aligning mechanism, including:

[0066] Step 21, the primary camera takes a picture of the crop rows and extracts the error angle as the moving angle θ, the method of calculating the error angle is shown in Figure 8 ;

[0067] Step 22, the primary row alignment mechanism pushes the parallel cantilever to align the center position line with the navigation line; the primary control is the lengthening / shortening amount ΔD1 of the hydraulic cylinder, ΔD1 = moving angle θ × cantilever length L, the cantilever length L is an input value, in this embodiment, it is the distance from the hinge point where the angle sensor is installed on the diamond-shaped cantilever 14 to the connection point between the hydraulic cylinder 15 and the diamond-shaped cantilever 14;

[0068] Step 3, initialize the secondary row spacing adjustment mechanism, including:

[0069] Step 31, the secondary camera takes a picture of 4 crop rows and fits 4 crop row lines, the method of extracting the crop row lines is shown in Figure 10 ;

[0070] Step 32, taking the position of the first column of corn rows as the zero point, according to the row line information, the position coordinate information of the second, third, and fourth columns of crop rows is obtained, and the difference between the position coordinate information and the position of the first column of corn rows is obtained. The intersection line of the inter-plant flexible weeding finger wheel in the second, third, and fourth columns is adjusted to align with the navigation line; the secondary control is the lengthening / shortening amount (ΔS2) of the electric push rod, ΔS2 = electric push rod speed v × time t;

[0071] Step 4, after the row spacing is determined, during the operation of the weeding machine, only the error angle is tracked by the primary camera to keep the center line aligned with the navigation line, and the primary row alignment mechanism is adjusted according to the error angle to offset the entire weeding unit when needed;

[0072] Before step 1, calibration is performed, taking the weeding center line as the position of the center of the weeding component, then according to the input spacing between each row weeding production combination 20 as the spacing between the actual position lines of each column of weeding components, and resetting all electric push rods 7, assuming that the crop row spacing is 700 mm, the initial position of the electric cylinder (in this example, the unit is mm) is 0, 700, 1400, 2100. If it is identified that the position of the row line at the bottom of the image is 210, 920, 1590, 2410. Then the positions to be moved by the four electric cylinders are ΔS1a = 210-200, ΔS1b = 920-700, ΔS1c = 1590-1400, ΔS1d = 2410-2100.

[0073] The method of extracting the error angle in steps 21 and 4 is shown in Figure 8 and Figure 9 , including:

[0074] Step 211. Input the image taken by the primary camera;

[0075] Step 212. Perform grayscale conversion, Gaussian blur and edge detection on the image captured by the primary camera in sequence;

[0076] Step 213. ROI acquisition and perspective transformation: acquire the region of interest (ROI) according to the width and height of the image and a preset ratio, and apply perspective transformation to the ROI to transform the image from the original view to the bird's eye view;

[0077] Step 214. Calculate the histogram and find the histogram peaks: calculate the histogram of the transformed image and find the peaks on the left and right sides, which correspond to the positions of the left and right sides of the crop row;

[0078] Step 215. Find the crop row using the sliding window method: use the histogram peaks as the starting point and apply the sliding window method upwards to find the pixel index of the entire crop row. Through these pixels, a polynomial curve representing the crop row can be fitted;

[0079] Step 216. Polynomial curve fitting and crop row: use the found pixel points to fit a polynomial curve to estimate the path of the crop row;

[0080] Step 217. Inverse perspective transformation: inverse perspective transform the bird's eye view containing the estimated crop row back to the original view;

[0081] Step 218. Calculate the evaluation line center point: at a certain height (near the bottom of the image), calculate the left and right boundary points of the crop row, and then calculate the center position of these two boundary points as the evaluation line center point, which represents the current position of the machine relative to the crop row;

[0082] Step 219. Calculate the target center point: at another higher position in the image, repeat the above process to calculate the left and right boundary points of the crop row and their center position, which represents the expected position of the machine at a future time point, that is, if the robot or vehicle maintains the current heading, it will reach this position;

[0083] Step 2110 Calculate the error angle: fit two straight lines from the evaluation line center points and the target center point: one is the current travel direction (from the center point at the bottom of the image to the center point at the top of the image, theoretically collinear with the weed removal center line on one side of the primary camera), and the other is the expected travel direction (ideally, this straight line should be parallel to the crop row and pass through the center of the image); the heading angle error is the included angle between the two straight lines;

[0084] The fitting method of the crop row line in step 31 is shown in Figure 10~Figure 12 , which includes:

[0085] Step 311, take a picture by the secondary camera;

[0086] Step 312, using Unet semantics to segment the image to extract the posture of each row of corn seedlings;

[0087] Step 313, converting into a gray image according to the semantic segmentation result;

[0088] Step 314, performing edge detection on the gray image after Gaussian blur processing to reduce noise;

[0089] Step 315, detecting the corner points (feature points) in the blurred image;

[0090] Step 316, drawing a vertical line as a crop row line according to the x coordinate determined by the feature points.

Claims

1. A precision on-row secondary regulation weeding machine, characterized in that, The utility model relates to a kind of inter-row and inter-plant weeding machines, including: main beam (19), first alignment mechanism (2), main frame (1), second row spacing adjusting mechanism (4) and inter-row and inter-plant weeding unit (3), wherein main beam (19) is fixed with the rear frame of tractor, the front and rear ends of first alignment mechanism (2) are fixed with main beam (19) and main frame (1) respectively, the front and rear ends of second row spacing adjusting mechanism (4) are fixed with main frame (1) and inter-row and inter-plant weeding unit (3) respectively;First camera (6) and second camera (5) are fixed at the left and right ends of main frame (1) respectively, first camera (6) is forwardly photographed, and second camera (5) is downwardly photographed;The number of inter-row and inter-plant weeding unit (3) is at least one group, and the number of second row spacing adjusting mechanism (4) is consistent with the number of inter-row and inter-plant weeding unit (3). The first alignment mechanism includes: rhombic cantilever (14), cantilever front mounting bracket (21), cantilever rear mounting bracket (22) and hydraulic cylinder, the front and rear ends of two rhombic cantilevers (14) are hinged with cantilever front mounting bracket and cantilever rear mounting bracket as the front and rear ends of the first alignment mechanism, and the connecting line of the front and rear hinge points of the rhombic cantilever (14) is parallel to the connecting line of the front and rear hinge points of the other rhombic cantilever (14);Any hinge point of two rhombic cantilevers (14) is provided with an angle sensor;One end of the hydraulic cylinder is fixed with main beam (19), and the other end of the hydraulic cylinder is connected with the rhombic cantilever (14) provided with the angle sensor, and the hydraulic cylinder is used to push the relative displacement between main beam (19) and main frame (1). The second row spacing adjusting mechanism (4) includes: electric push rod (7), pulley mounting plate (8), connecting rod (9) and weeding unit fixing plate (10), wherein the shell of the electric push rod is fixed with main frame (1) as the front end of the second row spacing adjusting mechanism (4), the movable end of the electric push rod (7) is fixed with the weeding unit fixing plate (10) through the connecting rod (9), two sections of the pulley mounting plate (8) are fixed with a weeding unit fixing plate (10) respectively, the pulley mounting plate (8) is provided with two groups of side shift pulleys (11) through deep groove ball bearings (12) and pulley mounting shafts (13) installed side by side, and the two groups of side shift pulleys (11) are in contact with main frame (1);The weeding unit fixing plate (10) is fixed with inter-row and inter-plant weeding unit (3) as the rear end of the second row spacing adjusting mechanism (4). The inter-row and inter-plant weeding unit (3) includes: weeding equipment rack (18), inter-row weeding production combination (20), inter-plant flexible weeding finger wheel (15), profiling mechanism (16) and press wheel (17), wherein the two ends of the profiling mechanism (16) are fixed with the weeding unit fixing plate (10) and the weeding equipment rack (18) respectively, the inter-row weeding production combination (20) and the inter-plant flexible weeding finger wheel (15) are fixed and installed below the weeding equipment rack (18), and the press wheel (17) is installed on the weeding equipment rack (18) and located in front of the inter-row weeding production combination (20).

2. The precision row-by-row secondary regulation weeding machine according to claim 1, characterized in that, The inter-row weeding production combination (20) includes two A-type soil separating shovels for removing inter-row weeds in crops.

3. The precision row-by-row secondary regulation weeding machine according to claim 2, characterized in that, ​ 4. The precision row-by-row secondary regulation weeding machine according to claim 1, characterized in that, The main frame (1) comprises two front and rear beams for mounting a first row alignment mechanism (2) and a second row spacing adjustment mechanism (4).

5. The method of claim 1, wherein the method is used for aligning a precision weeding machine with a secondary control. Comprise: Step 1, reset, initially, the first row alignment mechanism and the second row spacing adjustment mechanism are in the initial position; Step 2, initialize the first row alignment mechanism, including: Step 21, the first camera shoots the crop row and extracts the error angle as the moving angle θ, Step 22, the hydraulic cylinder in the first row alignment mechanism pushes the parallel cantilever to make the center line align with the navigation line; the first control is the lengthening / shrinking amount ΔD1 of the hydraulic cylinder, ΔD1 = moving angle θ × cantilever length L, and the cantilever length L is an input value; Step 3, initialize the second row spacing adjustment mechanism, including: Step 31, the second camera shoots the crop row and fits four crop row lines, Step 32, taking the position of the first corn row as the zero point, the position coordinate information of the second, third and fourth corn rows is obtained, and the difference between the position of the first corn row and the position of the second, third and fourth corn rows is obtained; the intersection line of the flexible weeding finger in the second, third and fourth corn rows is adjusted to align with the navigation line; the second control is the lengthening / shrinking amount ΔS2 of the electric push rod, ΔS2 = electric push rod speed v × time t; Step 4, after the row spacing is determined, during the operation of the weeding machine, the error angle is tracked by the first camera to keep the center line aligned with the navigation line.

6. The alignment method of claim 5, wherein, Before step 1, calibration is performed, the weeding center line is defined as the position of the center of the weeding component, and then the spacing between the input various row weeding assemblies is taken as the spacing between the actual position lines of the various row weeding components, and all the electric push rods are reset.

7. The alignment method of claim 5, wherein the alignment method further comprises: The method for extracting the error angle in step 21 and step 4, comprising: Step 211. input the image shot by the first camera; Step 212. perform grayscale conversion, Gaussian blur and edge detection on the image shot by the first camera in sequence; Step 213. ROI acquisition and perspective change: acquire the region of interest (ROI) according to the width and height of the image and a preset ratio, and apply perspective transformation to the ROI, so as to change the image from the original view to the bird's eye view; Step 214. calculate the histogram and find the histogram peak value: calculate the histogram of the transformed image, and find the peak values on the left and right sides, which correspond to the positions on the left and right sides of the crop row; Step 215. find the crop row using the sliding window method: use the histogram peak value as the starting point, and apply the sliding window method upward to find the pixel index of the entire crop row, and through these pixels, a polynomial curve representing the crop row can be fitted; Step 216. polynomial curve fitting and crop row: use the found pixel points to fit a polynomial curve to estimate the path of the crop row; Step 217. inverse perspective transformation: inverse perspective transform the bird's eye view containing the estimated crop row back to the original view; Step 218. calculate the evaluation line center point: calculate the left and right boundary points of the crop row at the height of the bottom of the image, and then calculate the center position of the two boundary points as the evaluation line center point, which represents the position of the current machine relative to the crop row; Step 219. Calculate target center point: repeat the above process to calculate the left and right boundary points of the crop row and the center position of the crop row at another higher position in the image, the center position of the left and right boundary points represents the position that the machine is expected to reach at a certain time in the future, that is, if the robot or vehicle maintains the current heading, it will reach this position; Step 2110 Calculate error angle: fit the center points of each evaluation line and the target center point to form two straight lines: one is the current direction of travel, the other is the expected direction of travel; the heading angle error is the included angle between the two straight lines.

8. The alignment method of claim 5, wherein, The fitting method of the crop row line in step 31 includes: Step 311, take a picture by the secondary camera; Step 312, use Unet semantics to segment the image to extract the posture of each row of corn seedlings; Step 313, convert the semantic segmentation result into a gray-scale image; Step 314, perform edge detection on the gray-scale image after Gaussian blur processing to reduce noise; Step 315, detect the corner points in the blurred image; Step 316, draw a vertical line as the crop row line according to the x-coordinate determined by the feature points.

Images