Method for measuring tilling depth of rotary cultivator for soil bin test

Abstract

The invention relates to a method for measuring the tilling depth of a rotary cultivator for a soil bin test, which comprises the following steps of: 1) arranging a plurality of cylindrical targets on two sides of a soil bin test bed, and arranging black and white stripe-shaped marks on the targets; (2) two cameras are arranged on a mud cover shell of the rotary cultivator; the camera is electrically connected with the industrial personal computer; meanwhile, performing joint calibration on the camera, and establishing a fitting model: W = f (L, H) = a1L + a2H + a3 or W = b1L2 + b2H2 + b3L + b4H + b5; 3) placing the rotary cultivator at a preset initial position of a soil bin test bed, and obtaining a reference height H0 of tilling depth measurement through the fitting model; (4) starting the rotary cultivator to carry out a soil bin cultivation test, synchronously tracking and shooting the marks on the two sides of the soil bin in real time by the camera, then obtaining two groups of side cultivation depths through the fitting model, namely h1 = H0-H1 and h2 = H0-H2, and (5) obtaining the actual cultivation depth h = k1h1 + k2h2 after redundancy verification is carried out on the side cultivation depths in the step (4). The method is convenient and fast to install, accurate in measurement, high in anti-interference capability and capable of carrying out real-time dynamic measurement.

Description

Technical Field

[0001] This invention relates to a method for identifying agricultural machinery, and more particularly to a method for measuring the tillage depth of a rotary tiller, specifically a method for measuring the tillage depth of a rotary tiller used in a soil trough test. Background Technology

[0002] Rotary tillers, as commonly used soil tillage machinery in agricultural production, have tillage depth as a key indicator for measuring operational quality and performance, and it is also one of the core testing items in agricultural machinery appraisal. The soil trench test is an important means of simulating the field working environment and conducting performance testing during agricultural machinery appraisal. Accurately measuring the tillage depth of the rotary tiller in the soil trench test is of great significance for objectively evaluating the performance of the rotary tiller and ensuring the quality of agricultural machinery products.

[0003] Existing methods for measuring tillage depth using rotary tillers mainly fall into two categories: mechanical contact measurement and traditional photoelectric measurement. Mechanical contact measurement typically uses tools such as depth gauges and poles, with data read manually. This method is not only cumbersome and inefficient, but the measurement results are also easily affected by human error, making real-time dynamic measurement difficult. Traditional photoelectric measurement methods often rely on a single sensor or camera, which is greatly affected by the soil test environment (such as soil dust and changes in light). The measurement accuracy is unstable, and there is a lack of data redundancy verification mechanisms. When a single measurement unit malfunctions, it can easily lead to measurement interruption or data loss.

[0004] In addition, the rotary tiller's working posture is complex in the soil trough test, and traditional measuring devices often have problems such as inconvenient installation and poor adaptability, which affect the accuracy of tillage depth measurement.

[0005] Therefore, improvements are urgently needed to meet the high-standard testing requirements in the field of agricultural machinery appraisal. Summary of the Invention

[0006] The purpose of this invention is to address the shortcomings of existing technologies by providing a method for measuring the tillage depth of rotary tillers in soil trough tests. This method is easy to install, accurate in measurement, has strong anti-interference capabilities, and can achieve real-time dynamic measurement of the tillage depth of rotary tillers, fully meeting the needs of agricultural machinery appraisal.

[0007] The technical solution of this invention is: A method for measuring the tillage depth of a rotary tiller used in a soil trough test includes the following steps: 1) On both sides of the soil test bench, several cylindrical targets are evenly arranged along its length, and black and white striped markings are set on the targets. 2) Two cameras are installed on the mudguard shell of the rotary tiller, facing the targets on both sides of the soil trough test platform. These cameras are electrically connected to an industrial control computer (ICC) to transmit the captured information and perform calculations. Simultaneously, the cameras are jointly calibrated to obtain their intrinsic and extrinsic parameters, and a fitting model is established: W=f(L,H)= a1L+a2H+a3 (linear fitting) or W=b1L 2 +b2H 2 +b3L+b4H+b5 (polynomial fitting), where a1, a2, a3, b1, b2, b3, b4, and b5 are fitting coefficients, L is the horizontal distance between the camera and the target; W is the pixel width of the stripes in the image; and H is the vertical height of the camera. 3) Place the rotary tiller in the preset initial position on the soil trough test platform, and obtain the reference height H0 for tillage depth measurement through the fitting model; 4) Start the rotary tiller to conduct a soil trench tillage test. At the same time, the camera tracks and photographs the markings on the targets on both sides of the soil trench in real time. Then, through the fitting model, calculate and obtain the real-time heights H1 and H2 of the cameras on both sides, and obtain two sets of side tillage depths: h1=H0-H1, h2=H0-H2. 5) After redundancy verification of the lateral tillage depth in step 4), obtain the actual tillage depth: When |h1-h2|≤ preset threshold, the measurement data is deemed valid, and the actual tillage depth h after fusion is output as h=k1h1+k2h2; where k1 and k2 are weighting coefficients, determined according to the calibration accuracy of the two cameras; When |h1-h2|> the preset threshold, an error message is issued, and the most recent historical data of the actual tillage depth h is used instead to ensure the continuity of the measurement.

[0008] Furthermore, the marking is a striped sticker with a white base and several horizontal black stripes printed on it; the marking is affixed to the surface of the target radially, and the stripes are perpendicular to the axis of the target.

[0009] Furthermore, the camera is a binocular stereo camera and is mounted on the mudguard housing of the rotary tiller via an electric gimbal.

[0010] Furthermore, the base of the electric gimbal is equipped with a magnetic structure for easy installation and disassembly.

[0011] Furthermore, the camera corresponds to the target located nearby.

[0012] Furthermore, the intrinsic parameters include focal length and pixel size; the extrinsic parameters are relative positional relationships.

[0013] Furthermore, in step 3), the camera photographs the target, and the pixel width W0 of the black stripes in the image is extracted by the machine vision algorithm; combined with the parallax principle of the binocular stereo camera, the horizontal distance L0 between the camera and the target is calculated in real time; then, W0 and L0 are substituted into the fitting model W=f(L,H) to obtain the vertical height H0 of the camera at this time, which is the reference height.

[0014] Furthermore, in step 3), H0 is the average value of the reference height obtained by taking pictures and calculating them from both cameras.

[0015] The beneficial effects of this invention are: 1. High measurement accuracy: This invention uses binocular vision technology combined with the stripe width recognition principle. It solves the camera's vertical height through a pre-established mathematical fitting model and uses two cameras to achieve data redundancy verification and fusion, effectively reducing environmental interference and measurement errors, and significantly improving the accuracy of tillage depth measurement.

[0016] 2. Easy installation and strong adaptability: The camera is magnetically mounted on the rotary tiller's mudguard, requiring no structural modifications to the tiller and making assembly and disassembly convenient. Furthermore, the radially attached stripes on the target ensure the camera can capture the stripes from any angle, adapting to changes in the rotary tiller's posture during operation.

[0017] 3. Real-time dynamic measurement: The entire measurement process requires no manual intervention and can capture the vertical displacement of the rotary tiller's mudguard in real time, quickly calculate and output tillage depth data, and meet the needs of dynamic operation performance testing of rotary tillers in agricultural machinery appraisal.

[0018] 4. High reliability: When one camera experiences a temporary malfunction or data anomaly, data from another camera can be used for redundancy verification, anomalies can be detected in a timely manner, and fitted values ​​can be used to replace them, thus improving the reliability of the measurement system. Attached Figure Description

[0019] Figure 1 This is a schematic diagram of the target of the present invention. Detailed Implementation

[0020] The present invention will be further described below with reference to the accompanying drawings and embodiments.

[0021] A method for measuring the tillage depth of a rotary tiller used in a soil trough test includes the following steps: Step 1: Target Placement Several cylindrical targets are evenly arranged along the length of both sides of the soil trough test platform, and striped markings are affixed to the targets, such as... Figure 1 As shown, the label is a striped sticker with several horizontal black stripes printed on a white background. The stripes are black rectangles, and their width and spacing are set according to experimental requirements.

[0022] Preferably, the label is affixed radially along the surface of the target to ensure that the camera can clearly capture black stripes of the same width when shooting from different angles at the same distance.

[0023] Step 2: Installation and Calibration of Measuring Device Two identical binocular stereo cameras were selected and mounted on two independent motorized pan-tilt units. The two motorized pan-tilt units were fixed to the mudguard shell of the rotary tiller via magnetic bases, ensuring that each camera was facing a nearby target with its imaging plane perpendicular to the ground. The cameras were electrically connected to an industrial control computer (ICC) to transmit the captured information and perform calculations.

[0024] Joint calibration of the two cameras was performed to obtain their intrinsic parameters, including focal length and pixel size, and extrinsic parameters, including relative positional relationships. Simultaneously, calibration experiments were conducted in an indoor environment: the cameras were fixed at different vertical heights and horizontal distances, and stripes on a target were photographed, with the pixel width of the stripes recorded in the 2D images. Then, a mathematical model of the relationship between stripe pixel width and the camera's vertical height and horizontal distance was established using linear or polynomial fitting methods, as follows: Let the horizontal distance between the camera and the target be L, the vertical height of the camera be H (with the bottom of the trench as the reference plane), and the pixel width of the stripes in the image be W. Then, establish the fitting model: W=f(L,H)= a1L+a2H+a3 (linear fitting) or W=b1L 2 +b2H 2 +b3L+b4H+b5 (polynomial fitting), where a1, a2, a3, b1, b2, b3, b4, and b5 are fitting coefficients, which are obtained by solving the least squares method using multiple sets of (L, H, W) data obtained in the calibration experiment.

[0025] Step 3: Calibration of reference position Place the rotary tiller in the preset initial position on the soil trough test platform, and start the rotary tiller to make the rotary blades rotate at a uniform speed.

[0026] Adjust the height of the rotary tiller so that the rotating envelope of the rotary blades is just tangent to the soil surface in the trench. At this point, the rotary blades are not yet in the soil, and the tillage depth is 0.

[0027] Two cameras are controlled to simultaneously track and photograph targets on both sides. The pixel width W0 of the black stripes in the images is extracted using a machine vision algorithm. The horizontal distance L0 between the camera and the target is calculated in real time using the camera parallax principle. W0 and L0 are substituted into the fitting model W=f(L,H) established in step 2 to obtain the vertical height H0 of the camera at this time. At this point, the rotary tiller blades have not yet entered the soil; therefore, this H0 is the reference height for tillage depth measurement. Preferably, the reference height values ​​obtained by photographing and calculating the targets from both cameras are averaged to obtain the final reference height, in order to reduce errors. The calculation of the horizontal distance based on the parallax principle is a mature existing technology and will not be described in detail here.

[0028] Step 4: Real-time tillage depth measurement Start the rotary tiller to conduct a soil trough tillage test, and allow the rotary tiller blades to enter the soil and operate normally. During the operation, two cameras are controlled to track and photograph the targets on both sides in real time, and a set of image data is acquired at preset intervals. Each set of image data is processed as follows: First, the horizontal distances L1 and L2 between the two cameras and the two targets are calculated using a binocular vision algorithm; then, the edges of the black stripes in the image are extracted using an edge detection algorithm, and the pixel widths W1 and W2 of the stripes are calculated. Substituting L1, W1 and L2, W2 into the fitting model established in step 2, respectively, the vertical heights H1 and H2 of the cameras on both sides are obtained by inverse solving. Then, the two sets of lateral tillage depths h1 and h2 are obtained as follows: h1 = H0 - H1, h2 = H0 - H2.

[0029] Step 5: Data Redundancy Verification and Output The measurement data is considered valid when |h1-h2| ≤ a preset threshold. Then, the fused actual tillage depth h is output as: h = k1h1 + k2h2, and this actual tillage depth h is stored. Here, k1 and k2 are weighting coefficients, determined based on the calibration accuracy of the two cameras.

[0030] When |h1-h2|>the preset threshold, the industrial control computer issues an error message and uses the most recent historical data of the actual tillage depth h to replace it, ensuring the continuity of the measurement.

[0031] The real-time tillage depth data is stored in the industrial control computer and displayed in real time on the screen, thus completing the tillage depth measurement in the rotary tiller trough test. Example

[0032] The soil trough test platform is 20m long, 1.5m wide, and 0.8m deep. Eight cylindrical targets with a diameter of 10cm and a height of 1.2m are selected and evenly arranged along both sides of the soil trough, with four targets on each side and a spacing of 5m between adjacent targets.

[0033] The striped sticker is made of white PVC material with a base width of 10cm and three horizontal black stripes printed on it. Each stripe is 2cm wide and the stripe spacing is 3cm.

[0034] Two stereo cameras with a resolution of 1920×1080 and a lens focal length of 8mm were selected. The two cameras were fixed in place using an aluminum alloy bracket, 50cm in length, ensuring that both cameras could capture the striped images of the targets on the left and right sides. The cameras were then fixed to the center of the rotary tiller's mudguard using a motorized pan-tilt head, with the left camera aligned with the target on the left side of the trough and the right camera aligned with the target on the right side of the trough.

[0035] The two cameras were jointly calibrated, and camera intrinsic parameters, including focal length f, were obtained using a checkerboard calibration board. x =1000 pixels, f y =1000 pixels, pixel size is 1.4μm×1.4μm; the relative distance between the two cameras in the external parameters is 50cm.

[0036] Calibration experiments were conducted: The camera was fixed at different combinations of horizontal distances L (1m, 1.5m, 2m, 2.5m, 3m) and vertical heights H (0.5m, 0.6m, 0.7m, 0.8m, 0.9m, 1.0m). Three images of the target were captured at each location, and the pixel width W of the stripes was measured. The average value was taken as the valid data for that location. A total of 30 sets of (L, H, W) data were obtained. Linear fitting was performed using the least squares method, resulting in the fitting model: W = 20L + 35H + 120. Here, W is in pixels, L is in meters, H is in meters, and the goodness of fit R is... 2 =0.992.

[0037] Reference position calibration: Start the rotary tiller and adjust its height so that the rotation envelope of the rotary blades is tangent to the soil surface of the trench. Control two cameras to photograph the targets on both sides, obtaining the stripe pixel width W. 0L =320 pixels, W 0R =318 pixels, horizontal distance L calculated by the binocular camera 0L =1.8m, L 0R =1.82m. Substituting the data into the fitting model, the inverse solution yields H. 0L =(320-20×1.8-120) / 35=0.8m, H 0R =(318-20×1.82-120) / 35=0.79m. For H 0L and H 0R Take the average value as the reference height H0 = (0.8 + 0.79) / 2 = 0.795m.

[0038] Real-time tillage depth measurement: The rotary tiller operates at a speed of 5 km / h, and the camera captures an image every 200 ms. At a certain moment, the left camera measures W. 1L =400 pixels, L 1L =1.8m, inverse solution yields H 1L =(400-20×1.8-120) / 35=0.52m. The right-side camera measured W. 1R =398 pixels, L 1R =1.82m, the inverse solution yields H 1R =(398-20×1.82-120) / 35=0.51m. Therefore, h1=0.795-0.52=0.275m, h2=0.795-0.51=0.285m.

[0039] Data validation and output: Set the threshold to 0.01m. Since |h1-h2|=0.01≤0.01m, the data is valid. Take k1=k2=0.5, then the actual tillage depth h=0.5×0.275+0.5×0.285=0.28m, and complete the output.

[0040] This invention features convenient installation, accurate measurement, and strong anti-interference capabilities. It can also achieve real-time dynamic measurement of the tillage depth of rotary tillers, which can fully meet the needs of agricultural machinery appraisal.

[0041] All parts not covered in this invention are the same as or can be implemented using existing technologies.

Claims

1. A method for measuring the tillage depth of a rotary tiller used in a soil trough test, characterized in that, Includes the following steps: 1) On both sides of the soil test bench, several cylindrical targets are evenly arranged along its length, and black and white striped markings are set on the targets. 2) Two cameras are installed on the mudguard shell of the rotary tiller, facing the targets on both sides of the soil trough test platform. These cameras are electrically connected to an industrial control computer (ICC) to transmit the captured information and perform calculations. Simultaneously, the cameras are jointly calibrated to obtain their intrinsic and extrinsic parameters, and a fitting model is established: W=f(L,H)=a1L+a2H+a3 (linear fitting) or W=b1L 2 +b2H 2 +b3L+b4H+b5 (polynomial fitting), where a1, a2, a3, b1, b2, b3, b4, and b5 are fitting coefficients, L is the horizontal distance between the camera and the target; W is the pixel width of the stripes in the image; and H is the vertical height of the camera. 3) Place the rotary tiller in the preset initial position on the soil trough test platform, and obtain the reference height H0 for tillage depth measurement through the fitting model; 4) Start the rotary tiller to conduct a trench tillage test. At the same time, the camera tracks and photographs the markings on the targets on both sides of the trench in real time. Then, through the fitting model, calculate and obtain the real-time heights H1 and H2 of the cameras on both sides, and obtain two sets of side tillage depths: h1=H0-H1, h2=H0-H2. 5) After redundancy verification of the lateral tillage depth in step 4), obtain the actual tillage depth: When |h1-h2|≤ preset threshold, the measurement data is deemed valid, and the actual tillage depth h after fusion is output as h=k1h1+k2h2; where k1 and k2 are weighting coefficients, determined according to the calibration accuracy of the two cameras; When |h1-h2|> the preset threshold, an error message is issued, and the most recent historical data of the actual tillage depth h is used instead to ensure the continuity of the measurement.

2. The method for measuring the tillage depth of a rotary tiller for a soil trough test according to claim 1, characterized in that, The marking is a striped sticker with a white base and several horizontal black stripes printed on it; the marking is affixed to the surface of the target radially, with the stripes perpendicular to the axis of the target.

3. The method for measuring the tillage depth of a rotary tiller for a soil trough test according to claim 1, characterized in that, The camera is a binocular stereo camera, and it is mounted on the mudguard shell of the rotary tiller via an electric gimbal.

4. The method for measuring the tillage depth of a rotary tiller for a soil trough test according to claim 3, characterized in that, The base of the electric gimbal is equipped with a magnetic structure for easy installation and disassembly.

5. The method for measuring the tillage depth of a rotary tiller for a soil trough test according to claim 1, characterized in that, The camera corresponds to the target located nearby.

6. The method for measuring the tillage depth of a rotary tiller for a soil trough test according to claim 1, characterized in that, The intrinsic parameters include focal length and pixel size; the extrinsic parameters are relative positional relationships.

7. The method for measuring the tillage depth of a rotary tiller for a soil trough test according to claim 1, characterized in that, In step 3), the camera takes a picture of the target and extracts the pixel width W0 of the black stripes in the image through a machine vision algorithm; combined with the parallax principle of the binocular stereo camera, the horizontal distance L0 between the camera and the target is calculated in real time; then, W0 and L0 are substituted into the fitting model W=f(L,H) to obtain the vertical height H0 of the camera at this time, which is the reference height.

8. The method for measuring the tillage depth of a rotary tiller for a soil trough test according to claim 1, characterized in that, In step 3), H0 is the average value of the reference height obtained by taking pictures and calculating the data from both cameras.

Images