Header component rotating speed automatic matching method and device, electronic equipment and harvesting machine

Abstract

The embodiment of the invention provides a header component rotating speed automatic matching method and device, electronic equipment and harvesting machinery, and relates to the technical field of agricultural machinery. The header component rotating speed automatic matching method comprises the steps that agronomic parameters of a to-be-picked area in front of a header are obtained; the agronomic parameters comprise at least one of size information of a to-be-picked object, distribution information of a stem where the to-be-picked object is located and the relative position of the to-be-picked object on the stem where the to-be-picked object is located; according to the agronomic parameters and the advancing speed of the header, the processing speed of the header for the to-be-picked object is obtained; and based on the processing speed of the header for the to-be-picked object, obtaining the operation speed of each processing part of the header for respectively executing the picking process. According to the embodiment, matching between the rotating speed of each part of the header and each monitoring parameter is achieved, and the working efficiency and the operation quality are improved.

Description

Technical Field

[0001] The present application relates to the technical field of agricultural machinery, and specifically to a method for automatically matching the rotation speed of a cutting table component, an automatic rotation speed matching device for a cutting table component, an electronic device, a corresponding storage medium, and a harvesting machine. Background Art

[0002] The improvement of the mechanization level of modern agriculture has put forward higher requirements for the performance and intelligence of agricultural machinery. Take corn harvesters as an example. As an important agricultural machinery and equipment, the working efficiency and operation quality of corn harvesters directly affect the harvesting effect of crops. At present, most corn harvesters on the market use a fixed speed header system. Although it can meet the basic harvesting needs, it is difficult for fixed speed headers to guarantee the best working efficiency and operation quality when facing different terrains, crop density and machine driving speed changes.

[0003] In the prior art, the harvesting machinery where the cutting table is located is provided with an engine electronic speed switch, which can correspond to multiple engine speeds. The operator can control the engine speed by rotating the button of the engine electronic speed switch, or by stepping on the accelerator. There is also a steering lever with a functional handle, which controls the displacement of the variable plunger pump by pushing the handle to control the vehicle speed. When the equipment is in operation, the operator operates the engine electronic speed switch to select different engine speeds. The engine speed will not be automatically adjusted after being set. The working parts of the cutting table are connected to the engine through a mechanical transmission mechanism. Due to the characteristics of the fixed transmission ratio of the mechanical transmission mechanism such as chain drive, gear drive and belt drive, the speed of the working parts of the cutting table can only be proportional to the engine speed and stable within a certain range.

[0004] In the prior art, since the working parts of the harvester, such as the stem pulling roller, the auger, and the straw chain, cannot adjust the speed according to different working conditions and requirements, the equipment may not be able to reach the optimal working state under certain working conditions. During operation, when the engine speed is maintained at a certain value, the speed of the harvester's working parts is also maintained at a certain value. It cannot be adaptively adjusted according to the driving speed of the corn machine and the agronomic conditions of the corn. However, the corn processing volume per second will fluctuate at different driving speeds. When the processing volume is small, power waste will occur. When the processing volume is large, the harvester will be blocked. Summary of the invention

[0005] The purpose of the embodiments of the present application is to provide a method, device, electronic device and harvesting machinery for automatic matching of the speed of the cutting table components. Based on key technologies such as machine vision, intelligent sensors, control algorithms and electronic control units, a method is provided for real-time monitoring of driving speed, cutting table height and various agronomic parameters during driving, and automatically adjusting the speed of the cutting table components through a hydraulic system, so as to at least solve some of the problems in the background technology.

[0006] In order to achieve the above object, a method for automatically matching the rotation speed of a header component is provided in the present application, and the method comprises: Acquire agronomic parameters of the area to be picked in front of the cutting table; the agronomic parameters include: size information of the objects to be picked, distribution information of the stems where the objects to be picked are located, and at least one of the relative positions of the objects to be picked on the stems where they are located; obtain the processing speed of the cutting table for the objects to be picked based on the agronomic parameters and the travel speed of the cutting table; and obtain the operating speeds of the various processing components of the cutting table used to respectively execute the picking process based on the processing speed of the cutting table for the objects to be picked.

[0007] Optionally, the processing speed of the cutting table for the objects to be picked is obtained based on the agronomic parameters and the travel speed of the cutting table, including: obtaining the plant processing speed based on the travel speed and the stem distribution information of the objects to be picked in the agronomic parameters; and converting the plant processing speed into the processing speed of the cutting table for the objects to be picked based on the number of objects to be picked per unit area.

[0008] Optionally, the processing components of the harvesting platform for executing the picking process include stem pulling rollers; based on the processing speed, the operating speeds of the various processing components of the harvesting platform for respectively executing the picking process are obtained, including: the stem pulling time is not greater than the driving time as the first constraint condition; the processing speed of the stem pulling roller to treat the picked objects is greater than the processing speed of the harvesting platform to treat the picked objects as the second constraint condition; the linear speed of the stem pulling roller that satisfies the first constraint condition and the second constraint condition is used as the minimum linear speed of the stem pulling roller.

[0009] Optionally, the first constraint condition is expressed as: ; in, V 拉 is the linear speed of the stem pulling roller, s is the safety factor, H 结穗 is the ear height, H 摘穗 is the cutting height of the header, t 车 is the time required for the header to travel one plant spacing; the second constraint condition is expressed as: ; Among them, Q 拉 The processing speed of the stem roller to be picked is n is the number of rows at harvest time, and Q is the processing speed of the harvesting platform.

[0010] Optionally, the processing components of the harvesting platform for executing the picking process include a straw chain; based on the processing speed, the operating speeds of the various processing components of the harvesting platform for respectively executing the picking process are obtained, including: obtaining an expression for the processing speed of the straw chain for the picked objects, the parameters of the expression including the linear speed of the straw chain; determining the value range of the linear speed of the straw chain with the constraint that the processing speed of the straw chain for the picked objects is greater than the processing speed of the harvesting platform for the picked objects; and obtaining the minimum linear speed of the straw chain based on the value range of the linear speed of the straw chain.

[0011] Optionally, the expression of the processing speed of the reed chain for the objects to be picked includes: ; Among them, Q 拨 The speed at which the harvesting chain processes the harvested material. n is the number of rows at harvest, Ψ 1 is the filling coefficient. Considering that the ears in the picking path are not completely full, the filling coefficient is between 0.5 and 1. L 0 is the length of corn ear, V 拨 is the linear speed of the reed chain.

[0012] Optionally, the processing components of the harvesting platform for executing the picking process include an auger; based on the processing speed, the operating speeds of the various processing components of the harvesting platform for respectively executing the picking process are obtained, including: obtaining an expression for the processing speed of the auger for the picked objects, wherein the parameters of the expression include the auger speed; determining the value range of the auger speed with the constraint that the processing speed of the auger for the picked objects is greater than the processing speed of the harvesting platform for the picked objects; and obtaining the minimum speed of the auger based on the value range of the auger speed.

[0013] Optionally, the expression for the processing speed of the auger on the harvested material comprises: ; Among them, Q 搅龙 To control the speed of the harvesting process, η is the actual efficiency value of the auger, P is the screw pitch of the auger, π is the circumference of a circle, d 1 is the outer diameter of the auger blade, d 2 is the inner diameter of the auger blade, V 1 is the volume of the material to be picked, n 搅龙 is the auger speed.

[0014] Optionally, agronomic parameters of the area to be harvested in front of the harvesting platform are obtained, including: obtaining a first image and a second image of the area to be harvested through a first camera and a second camera respectively; identifying a number of feature points of the objects to be harvested and the stems on which they are located from the first image and the second image; determining depth information of the number of feature points according to position parameters and optical parameters of the first camera and the second camera; determining a target to be measured based on the agronomic parameters to be determined, and selecting feature points of the target to be measured from the number of feature points; and obtaining the value of the agronomic parameter to be determined based on the depth information of the feature points of the target to be measured and the pixel size information of the image of the target to be measured.

[0015] Optionally, several feature points of the object to be picked and the stem on which it is located are identified from the first image and the second image, including: using a trained real-time target detection algorithm to identify the object to be picked and the stem on which it is located from the first image and the second image respectively; using a scale-invariant feature transformation algorithm to identify a feature point set from the object to be picked and the stem on which it is located, to obtain a first feature point set corresponding to the first image and a second feature point set corresponding to the second image; matching the first feature point set with the second feature point set to obtain several feature points of the object to be picked and the stem on which it is located.

[0016] Optionally, the real-time target detection algorithm is obtained by improving the YOLO algorithm as follows: introducing a channel-priority convolutional attention module into the backbone layer of the YOLO algorithm; and introducing a multi-scale convolutional module into the connection layer of the YOLO algorithm.

[0017] Optionally, the depth information of the plurality of feature points is determined according to the position parameters and optical parameters of the first camera and the second camera, including: for each of the plurality of feature points, at least one set of similar triangles is constructed by selecting a plurality of points from among the feature points, the feature point, the optical center of the first camera, the optical center of the second camera, the imaging point of the feature point in the first camera, and the imaging point of the feature point in the second camera; based on the similar triangles, a distance between the optical centers, a focal length of the first camera, a focal length of the second camera, and a distance between the imaging point and an end of the imaging surface is calculated to obtain a distance between the feature point and a line connecting the optical centers of the cameras or a distance between the feature point and the imaging surface; and the distance between the feature point and the line connecting the optical centers of the cameras or the distance between the feature point and the imaging surface is used as the depth information of the feature point.

[0018] Optionally, the target to be measured is determined based on the agronomic parameters to be determined, and the feature points of the target to be measured are selected from the several feature points, including: if the agronomic parameter to be determined is the size information of the object to be picked, the target to be measured is the object to be picked, and the feature points of the target to be measured include the left feature point, the right feature point, the top feature point and the bottom feature point of the object to be picked; if the agronomic parameter to be determined is the distribution information of the stem where the object to be picked is located, the target to be measured is the stem where the object to be picked is located, and the feature points of the target to be measured include the bottom feature point of the stem where the object to be picked is located; if the agronomic parameter to be determined is the relative position of the object to be picked on the stem where it is located, the target to be measured is the object to be picked and the stem where it is located, and the feature points of the target to be measured include the bottom feature point of the picked object and the bottom feature point of the stem where the object to be picked is located.

[0019] Optionally, the value of the agronomic parameter to be determined is obtained based on the depth information of the feature points of the target to be measured and the pixel size information of the image of the target to be measured, including: obtaining the pixel distance between feature points based on the pixel size information of the line segment formed by the feature points of the target to be measured in the image; obtaining the projection height of the line segment formed by the feature points of the target to be measured based on the pixel distance and the depth information, and obtaining the value of the agronomic parameter to be determined by the projection height.

[0020] Optionally, when the depth information of different feature points of the target to be measured is inconsistent, the method further includes: calculating the depth difference between different feature points in the same line segment, correcting the projection height based on the depth difference, and obtaining the value of the agronomic parameter to be determined by using the corrected projection height.

[0021] Optionally, after obtaining the operating speeds of each processing component used to respectively execute the picking process, the method further includes: adjusting the output flow of the hydraulic pump through a control signal, controlling the hydraulic motor to drive each processing component to operate at the operating speed; and obtaining the real-time speed of each working component of the cutting platform based on the current engine speed and feedback signal.

[0022] The present application also provides an automatic matching device for the rotation speed of cutting table components, which includes: a parameter acquisition module, used to acquire agronomic parameters of the area to be picked in front of the cutting table; the agronomic parameters include: size information of the objects to be picked, distribution information of the stems where the objects to be picked are located, and at least one of the relative positions of the objects to be picked on the stems where they are located; a processing speed module, which obtains the processing speed of the cutting table for the objects to be picked according to the agronomic parameters and the moving speed of the cutting table; and a speed output module, which is used to obtain the operating speeds of each processing component of the cutting table used to respectively execute the picking process based on the processing speed of the cutting table for the objects to be picked.

[0023] The present application also provides an electronic device, comprising: at least one processor; a memory connected to the at least one processor; wherein the memory stores instructions that can be executed by the at least one processor, and the at least one processor implements the aforementioned method for automatic matching of the rotation speed of the cutting table components by executing the instructions stored in the memory.

[0024] The present application also provides a machine-readable storage medium, on which instructions are stored. When the instructions are executed by a processor, the processor is configured to execute the aforementioned method for automatically matching the rotation speed of the header components.

[0025] The present application also provides a computer program product, including a computer program, which implements the aforementioned method for automatically matching the rotation speed of the header components when executed by a processor.

[0026] The present application also provides a harvesting machine, in which a PLC is configured to execute the aforementioned method for automatically matching the rotation speed of the cutting table components.

[0027] The above technical solution has the following beneficial effects: (1) Improve work efficiency and ensure operation quality. By detecting the working status and actual working conditions of the header, the speed of the header components is automatically adjusted, which improves work efficiency, prevents abnormal blockage of the header, reduces missed harvests and damage to crops, and improves operation quality.

[0028] (2) Reduce manual intervention. Automatic control systems reduce manual intervention and improve the convenience and accuracy of operations.

[0029] (3) Adaptability to different conditions. The system used to implement the method of the present application can maintain the best operating state under different terrains and working conditions and has strong adaptability.

[0030] Other features and advantages of the embodiments of the present application will be described in detail in the subsequent specific implementation section. BRIEF DESCRIPTION OF THE DRAWINGS

[0031] The accompanying drawings are used to provide a further understanding of the embodiments of the present application and constitute a part of the specification. Together with the following specific implementations, they are used to explain the embodiments of the present application, but do not constitute a limitation on the embodiments of the present application. In the accompanying drawings: Figure 1 A schematic diagram of the steps of the method for automatically matching the rotation speed of the header components according to an embodiment of the present application is schematically shown; Figure 2 The schematic diagram schematically shows the principle of a multi-scale convolution module according to an embodiment of the present application; Figure 3The schematic diagram of the model structure of the real-time target detection algorithm according to the implementation mode of the present application is shown; Figure 4 A flowchart based on the SIFT algorithm according to an embodiment of the present application is schematically shown; Figure 5 The schematic diagram of the binocular vision ranging model according to the embodiment of the present application is schematically shown; Figure 6 The schematic diagram schematically shows the principle of pinhole imaging according to the embodiment of the present application; Figure 7 A schematic diagram of the structure of the automatic speed matching device for the header components according to the embodiment of the present application is shown; Figure 8 The internal structure of the electronic device according to the embodiment of the present application is schematically shown; Fig. 9 A schematic diagram of the system structure of a harvesting machine according to an embodiment of the present application is shown schematically. DETAILED DESCRIPTION

[0032] The specific implementation of the embodiment of the present application is described in detail below in conjunction with the accompanying drawings. It should be understood that the specific implementation described here is only used to illustrate and explain the embodiment of the present application, and is not used to limit the embodiment of the present application.

[0033] Figure 1 The following is a schematic diagram showing the steps of the method for automatically matching the rotation speed of the header components according to the embodiment of the present application. Figure 1 As shown, a method for automatically matching the rotation speed of a header component, the method comprising: S01, obtaining agronomic parameters of the area to be picked in front of the header; the agronomic parameters include: at least one of the size information of the object to be picked, the distribution information of the stem where the object to be picked is located, and the relative position of the object to be picked on the stem where it is located; S02, obtaining a processing speed of the harvesting table for the harvested objects according to the agronomic parameters and the travel speed of the harvesting table; S03. Based on the processing speed of the harvesting platform on the objects to be harvested, the operating speeds of the respective processing components of the harvesting platform for respectively executing the harvesting process are obtained.

[0034] Through the above implementation method, the agronomic parameters of the harvested objects are identified, and the required rotation speed of each component of the harvesting table is calculated through the agronomic parameters, and then the rotation speed of the harvesting table components is adjusted through the hydraulic system to ensure that the machine works in the best condition and improve the overall operating efficiency.

[0035] For example, the object to be picked is corn, and the stalk where the object to be picked is located is the corn stalk where the corn is located. The subsequent implementation methods of this application are described using corn as an example, and other crops with similar picking methods to corn can also be used as a reference to implement the method in this application.

[0036] In some embodiments of the present application, the agronomic parameters of the area to be harvested in front of the header can be obtained through a visual recognition unit, the visual recognition unit includes a camera and a deep learning target detection model, the camera uses two array cameras to form a binocular camera to take pictures, the target detection model is trained so that it can accurately identify corn ears and corn stalks, and is embedded in the intelligent control unit. Specifically, the first image and the second image of the area to be harvested are obtained by the first camera and the second camera respectively; the first camera and the second camera can be two cameras in the binocular camera.

[0037] In some embodiments, two area array cameras can be installed on the lower side of the harvesting platform cab. The two cameras are located in the same horizontal plane with a horizontal distance of X1. The area array camera is controlled to take a picture every 1 second of driving. The harvesting platform width of this corn harvester model is X2, and its detection field of view is approximately a square area of ​​X1*X2, so as to detect the area to be harvested or picked.

[0038] Every time the header travels for 1 second, the subsequent deep learning algorithm and binocular vision algorithm are used to detect the agronomic parameters of corn in the detection area. The agronomic parameters are used to perform statistical analysis and calculate the required rotation speed of each component of the header, and then perform matching adjustment. Several feature points of the object to be picked and the stem where it is located are identified from the first image and the second image; the feature points are mainly points that reflect the position characteristics and shape characteristics of the object to be picked and the stem where it is located. The depth information of the several feature points is determined according to the position parameters and optical parameters of the first camera and the second camera; since the camera imaging is planar and the area to be picked is three-dimensional, the depth information of the feature points is required for subsequent calculations of size, length, distance, etc. The target to be measured is determined based on the agronomic parameter to be determined, and the feature points of the target to be measured are selected from the several feature points; since different agronomic parameters represent different meanings, different feature points need to be selected to calculate different agronomic parameters. The value of the agronomic parameter to be determined is obtained based on the depth information of the feature points of the target to be measured and the pixel size information of the image of the target to be measured. Based on the imaging distance between the feature points, the actual distance represented by the feature points in the area to be harvested is obtained, and finally the value of the required agronomic parameter is obtained.

[0039] In some embodiments of the present application, a real-time target detection algorithm YOLOv8 (YouOnly Look Once) based on deep learning is used to identify corn ears and corn stalks, and the algorithm is improved to a certain extent according to the characteristics of corn ears and corn stalks in terms of recognition and detection effect, so that it can accurately identify corn ears and corn stalks and mark them in the picture. The recognition accuracy can reach 99.5%, and the trained model is embedded in the intelligent control unit of the corn harvester to achieve end-to-end real-time detection.

[0040] The training process of the real-time target detection algorithm is as follows: in actual work, the corn field during the corn harvesting process is imaged, and the collected pictures are expanded. The main methods are image rotation, changing image brightness, adding Gaussian noise, filling rectangular areas, etc. The pictures are expanded to 2000, and the corn ears and corn stalks in the pictures are annotated by annotation software to form a data set. After annotation, a one-to-one annotation file is generated according to the annotated picture. The file contains the size, coordinates, category and other information of the corn ears and corn stalk targets to facilitate the model to perform recognition training. Finally, it is divided into a training set, a validation set and a test set according to a ratio of 7:2:1. This embodiment also designs a target detection network model training algorithm, including: on the training set, the improved model is trained with parameters, and the number of training times is 100. During the training process, the training parameters are verified by the validation set each time the training is performed, and the model parameters are corrected by the loss function and the gradient descent algorithm. This is repeated 100 times to obtain the final trained model, and the training results are tested with the test set to verify the accuracy of the model. The resulting real-time object detection algorithm can achieve accurate recognition of corn ears and corn stalks.

[0041] The real-time target detection algorithm adopted by the present invention is the improved YOLOv8. In order to further improve the recognition accuracy of the YOLOv8 algorithm on corn ears and corn stalks, and to realize real-time detection during the driving process of the corn machine and speed up its recognition speed, this embodiment proposes two innovative improvement measures to replace the modules in YOLOv8, so that it has better performance for corn stalks and corn ears.

[0042] Measure 1: Introduce a multi-scale convolution module in the connection layer of the YOLO algorithm. Figure 2 The schematic diagram of the principle of the multi-scale convolution module according to the implementation mode of the present application is shown schematically. Figure 2As shown in the figure, the basic principle of the multi-scale convolution module (MSConv) is to divide the channels of the input feature map into three parts according to the ratio of 1 / 2, 1 / 4, and 1 / 4, and perform standard convolution with convolution kernels of 1×1, 3×3, and 5×5 respectively. Then, all the convolved feature maps are spliced ​​together in the channel direction, and then a 1×1 convolution kernel is used for point-by-point convolution for further feature fusion. This convolution method enables it to generate more feature maps that can mine the required information from the features of corn ears and corn stalks at a lower cost, reduce the parameters of the model, and increase the model's ability to mine feature information, so that the model has higher accuracy and faster recognition speed, which is more in line with the requirements of real-time detection during the driving process of corn harvesters.

[0043] Measure 2: Introduce a channel-priority convolutional attention module in the backbone layer of the YOLO algorithm. The channel-priority convolutional attention mechanism (CPCA) is introduced. The CPCA model supports dynamic distribution of attention weights in the channel and spatial dimensions. Its overall structure includes the sequential arrangement of channel attention and spatial attention. The channel information of the feature map is aggregated by average pooling and maximum pooling of the channel attention. Then, the channel information is processed by a shared multi-layer perceptron (MLP) to generate a channel attention map. The input features are element-wise multiplied with the channel attention map to obtain the channel prior. The channel prior is then input into the deep convolution module to generate the spatial attention map. The deep convolution module receives the spatial attention map and performs channel mixing. Finally, the refined features are obtained as output by element-wise multiplication of the channel mixing result and the channel prior. Figure 3 The following schematically shows a model structure diagram of a real-time target detection algorithm according to an embodiment of the present application. Figure 3 As shown in the figure, the real-time target detection algorithm includes a backbone layer (Backbone), a connection layer (Neck) and a head layer (Head). MSConv is set in the connection layer, and the CPCA module is set after the SPPF (Spatial Pyramid Pooling – Fast) module in the backbone layer. The addition of the CPCA mechanism allows the model to focus on the features of corn ears and corn stalks with high weights when extracting features, while ignoring other irrelevant features in the image, making the model more accurate in detection and more adaptable to more complex working conditions of corn.

[0044] In some embodiments of the present application, several feature points of the object to be picked and the stem on which it is located are identified from the first image and the second image, including: using a trained real-time target detection algorithm to identify the object to be picked and the stem on which it is located from the first image and the second image respectively; using a scale-invariant feature conversion algorithm to identify a feature point set from the object to be picked and the stem on which it is located, to obtain a first feature point set corresponding to the first image and a second feature point set corresponding to the second image; matching the first feature point set with the second feature point set to obtain several feature points of the object to be picked and the stem on which it is located. For example: after accurately identifying and labeling the corn cobs and corn stalks in the picture, it is necessary to match the feature points of the corn cobs and corn stalks in the labeling box. In the subsequent agronomic calculation process, only the area containing 2*2 corn stalks is selected to represent the entire detection area, so it is only necessary to match the feature points in the labeling box of this area.

[0045] Figure 4 The flowchart based on SIFT algorithm according to the implementation mode of the present application is schematically shown. The SIFT (Scale-invariant feature transform) algorithm is used to detect the features of corn ears and corn stalks in the annotation box, and the feature points on the left and right views of corn ears and corn stalks taken by two cameras are matched in combination with the Euclidean distance constraint. To detect feature points in an image using the SIFT algorithm, it is first necessary to construct a Gaussian pyramid, which includes the following steps: Step 1: First, the image is enlarged by doubling it as the first layer of the first group of the Gaussian pyramid, and then Gaussian convolution is performed in sequence to obtain a total of 6 layers of images. The Gaussian convolution process is expressed as: , Among them, G(x,y,σ) is the Gaussian convolution function, I(x,y) is the value of each pixel in the image, and σ is a fixed parameter value.

[0046] Step 2: Downsample the third-to-last layer image of the first group as the first layer of the second group. That is, for each row, take every other pixel to get an image with rows and columns that are 1 / 2 of the original image.

[0047] Step 3: Repeat Step 1 to get the second set of images.

[0048] Step 4: Repeat Step 2 to get O groups of images, a total of O×6 images are the Gaussian pyramid. The images in adjacent groups can be subtracted to get the Gaussian difference pyramid (DOG). The feature points are composed of local extreme points found in the DOG space, so in order to detect the extreme points in the DOG, each pixel needs to be compared with all the adjacent points around it to get the feature points in the image.

[0049] After obtaining the feature points, its feature information needs to be described as follows: Step 1: First, extract the distribution characteristics of the gradient m(x, y) and direction θ(x, y) of each pixel in the 3×3 neighborhood of the Gaussian pyramid image where the feature point is located. The formula is as follows: , Step 2: After the acquisition is completed, the gradient and direction of all pixels are counted using a histogram. The histogram divides the direction of 0-360° into 8 columns, each of which is 45°. In the histogram, the peak direction is regarded as the main direction of the feature point. To improve robustness, directions higher than 80% of the main direction will be retained as auxiliary directions.

[0050] Step 3: After the above process, for each feature point, there are three pieces of information: position, scale and direction. In order to complete the feature point matching in the two images, a descriptor should be constructed for each feature point information, that is, the feature point information is vectorized. : First, determine the image area required to calculate the descriptor based on the Gaussian image of the scale of the feature point. The radius calculation formula is as follows: , Among them, d=4, which means that the image area is divided into 4×4 sub-blocks.

[0051] Step 4: Rotate the coordinate axis of the feature point to the direction of the feature point. Perform histogram statistics in 8 directions for each of the 4×4 sub-blocks to obtain the gradient amplitude in each direction. In this way, a 128-dimensional feature point description vector is constructed for each feature point.

[0052] The Euclidean distance between the feature points in the two images is calculated by the feature vector. The calculation formula is as follows: , Among them, a(i) is the feature vector of the feature points in a specific area of the corn ear and corn stalk images captured by the left camera, and b(i) is the feature vector of the feature points in a specific area of the corn ear and corn stalk images captured by the right camera. A threshold T is set, and when d < T, the feature point matching is successful. Through the above feature point matching, the first feature point set corresponding to the first image and the second feature point set corresponding to the second image can be aligned for subsequent calculations.

[0053] In some embodiments of the present application, determining the depth information of the several feature points according to the position parameters and optical parameters of the first camera and the second camera includes: for one of the several feature points, selecting several points from the optical center of the first camera, the optical center of the second camera, the feature point, the imaging point of the feature point on the first camera, and the imaging point of the feature point on the second camera to construct at least one set of similar triangles; based on the similar triangles, using several line lengths among the distance between the optical centers, the focal length of the first camera, the focal length of the second camera, and the distance from the imaging point to the end of the imaging plane, calculating the distance between the feature point and the connection line of the camera optical center or the distance between the feature point and the imaging plane; using the distance between the feature point and the connection line of the camera optical center or the distance between the feature point and the imaging plane as the depth information of the feature point.

[0054] Figure 5 Schematically shows a schematic diagram of the binocular vision ranging model according to an embodiment of the present application. As Figure 5 shown, in the ideal binocular vision ranging model diagram, the left and right cameras installed under the cab should have the same camera parameters. O L and O R respectively represent the optical centers of the left and right cameras. The optical axes of the two cameras are parallel to each other and are located on the same horizontal plane. The horizontal distance between the two optical centers is b. The feature point P is imaged as point P L and point P R on the phase planes of the left and right cameras respectively. The straight line where points P L and P R are located is parallel to the connection line of the camera optical centers O R O L . The distance from point P L to the leftmost end of its imaging plane is X L , and the distance from point P R to the leftmost end of its imaging plane is X R .

[0055] According to the principle of similar triangles, it can be deduced that: , , Adding the two equations gives: , Further deduction yields: , According to the position difference of the feature point at the bottom of the meter pole on the X-axis of the left and right camera image planes, the distance Z from this feature point to the camera optical center plane can be calculated according to the horizontal distance between the two cameras and the focal length f of the camera. Similarly, the depth information of the feature point at the bottom of the corn ear can also be obtained.

[0056] from Figure 5 It can also be seen that the figure contains multiple groups of similar triangles, and those skilled in the art can calculate the above calculation method by using different similar triangles as needed. This application does not list them one by one here.

[0057] In some embodiments of the present application, the target to be measured is determined based on the agronomic parameters to be determined, and the feature points of the target to be measured are selected from the several feature points, including: if the agronomic parameter to be determined is the size information of the object to be picked, the target to be measured is the object to be picked, and the feature points of the target to be measured include the left feature point, the right feature point, the top feature point and the bottom feature point of the object to be picked; if the agronomic parameter to be determined is the distribution information of the stem where the object to be picked is located, the target to be measured is the stem where the object to be picked is located, and the feature points of the target to be measured include the bottom feature point of the stem where the object to be picked is located; if the agronomic parameter to be determined is the relative position of the object to be picked on the stem where it is located, the target to be measured is the object to be picked and the stem where it is located, and the feature points of the target to be measured include the bottom feature point of the picked object and the bottom feature point of the stem where the object to be picked is located. For example, after the feature point matching is completed, the parameter values ​​of corn plant spacing, corn cob length, diameter, ear height and so on are calculated. The corn plant spacing can be obtained based on the front and rear feature points at the bottom of the corn stalk, the corn cob length can be obtained based on the feature points at the top and bottom of the corn cob, the corn cob diameter can be obtained based on the feature points at the left and right ends of the corn cob, and the ear height can be obtained based on the feature points at the bottom of the corn stalk and the bottom of the corn cob.

[0058] In some embodiments of the present application, the value of the agronomic parameter to be determined is obtained based on the depth information of the feature points of the target to be measured and the pixel size information of the image of the target to be measured, including: obtaining the pixel distance between the feature points based on the pixel size information of the line segment formed by the feature points of the target to be measured in the image; obtaining the projection height of the line segment formed by the feature points of the target to be measured based on the pixel distance and the depth information, and obtaining the value of the agronomic parameter to be determined by the projection height. Figure 6 The schematic diagram of the principle of pinhole imaging according to the embodiment of the present application is schematically shown. Figure 6As shown in the figure, a method for measuring the ear height is designed based on its depth information and the principle of pinhole imaging. First, the height of the corn ear height projected onto a plane parallel to the phase plane is calculated based on the feature points at the bottom of the corn stalk and the bottom of the corn ear, which can be expressed as: , where L is the projection height of the corn ear height, l is the pixel length corresponding to the ear height in the pixel coordinate after camera imaging, dx is the size occupied by a single pixel in the camera, f is the camera focal length, D It is the depth value from the horizontal plane where the target line segment is located to the optical center of the camera.

[0059] Furthermore, when the depth information of different feature points of the target to be measured is inconsistent, the method further includes: when the depth information of different feature points of the target to be measured is inconsistent, the method further includes: calculating the depth difference between different feature points in the same line segment, correcting the projection height based on the depth difference, and obtaining the value of the agronomic parameter to be determined with the corrected projection height. Since the growth direction of corn during the harvesting process is not completely perpendicular to the camera shooting direction, the entire corn stalk cannot be completely parallel to the camera's phase plane when the camera is shooting, so it is necessary to correct the projection height in combination with the depth difference. The correction method is exemplified as follows: The ear height can be calculated by combining the depth difference h of the feature point at the bottom end of the corn stalk and the feature point at the bottom end of the corn ear and the projection height L. H 结穗 : ; Similarly, the plant spacing, corn cob length and diameter can be calculated using the above method.

[0060] In some embodiments of the present application, the processing speed of the harvesting platform for the objects to be picked is obtained according to the agronomic parameters and the travel speed of the harvesting platform, including: obtaining the plant processing speed according to the travel speed and the distribution information of the stems of the objects to be picked in the agronomic parameters; converting the plant processing speed to the processing speed of the harvesting platform for the objects to be picked according to the number of objects to be picked per unit area. For example, after obtaining the agronomic parameters of corn, the total number of ears Q harvested by the harvesting platform in the next second can be obtained according to the plant spacing and the travel speed detected by the speed sensor, which is usually consistent with the driving speed of the harvesting machine where the harvesting platform is located, that is, equal to the driving speed of the corn machine. The calculation formula is: ; Where n is the number of rows at harvest; V 车 is the travel speed in m / s; L 株距 is the corn plant spacing, in mm; Q is the total number of ears harvested in the next second, in ear / s.

[0061] In some embodiments of the present application, the processing components of the harvesting platform for performing the picking process include a stem pulling roller; and the operation speeds of the respective processing components of the harvesting platform for performing the picking process respectively are obtained based on the processing speed, including: The first constraint condition is that the stem pulling time is not greater than the driving time, that is: ; in, V 拉 is the linear speed of the stem pulling roller, s is the safety factor, H 结穗 is the ear height, H 摘穗 is the cutting height of the header, t 车 The time required for the header to travel one plant spacing; in: , L 株距 is the corn plant spacing, V 车 For the travel speed.

[0062] The second constraint condition is that the processing speed of the stem pulling roller for picking is greater than the processing speed of the cutting table for picking, that is: ; Among them, Q 拉 The processing speed of the stem roller to be picked is n is the number of rows at harvest, Q is the processing speed of the harvesting platform; After transforming the above formula, we get: ; According to the above formula, the minimum linear speed required for the stem pulling roller can be determined according to the current travel speed, header ear picking height, corn ear height, and corn plant spacing. V 拉 ,Should V 拉 The first constraint and the second constraint can be satisfied.

[0063] In some embodiments of the present application, the processing components of the harvesting platform for performing the picking process include a reed chain; based on the processing speed, the operation speeds of the various processing components of the harvesting platform for respectively performing the picking process are obtained, including: obtaining an expression of the processing speed of the reed chain for the objects to be picked, the parameters of the expression including the linear speed of the reed chain; the expression is constructed according to the processing capacity of the picking path for the corn ears, which includes a variety of factors affecting the processing capacity, such as the filling coefficient, the holding capacity, the length of the corn ear and the linear speed of the reed chain, and the following is obtained: ; Among them, Q 拨The speed at which the harvesting chain processes the harvested material. n is the number of rows at harvest, Ψ 1 is the filling coefficient. Considering that the ears in the picking path are not completely full, the filling coefficient is between 0.5 and 1. L 0 is the length of corn ear, V 拨 is the linear speed of the reed chain.

[0064] The conveying capacity of the straw chain is Q 拨 It should be greater than the total number of ears Q harvested by the harvesting platform in the next second to ensure that the processing of the plants is completed. Therefore, the processing speed of the harvesting chain for the harvested objects is greater than the processing speed of the harvesting platform for the harvested objects as a constraint, that is: ; Determine the value range of the linear speed of the reed chain, namely: ; The minimum linear speed of the reel chain is obtained based on the value range of the linear speed of the reel chain.

[0065] In some embodiments of the present application, the processing components of the harvesting platform for performing the picking process include an auger; based on the processing speed, the operation speed of each processing component of the harvesting platform for performing the picking process respectively is obtained, including: obtaining an expression of the processing speed of the auger for the picked object, the expression is constructed according to the processing capacity of the auger for the corn cob, and includes multiple factors that affect the processing capacity, such as the auger blade size, the auger pitch, the actual efficiency of the auger, and the auger speed, and obtaining: ; Among them, Q 搅龙 To control the speed of the harvesting process, η is the actual efficiency value of the auger, P is the screw pitch of the auger, π is the circumference of a circle, d 1 is the outer diameter of the auger blade, d 2 is the inner diameter of the auger blade, n 搅龙 is the auger speed, V 1 is the volume of the object to be picked, expressed as: , D 0 is the diameter of the corn ear, L 0 is the length of the corn ear.

[0066] The auger needs to transport all the fruit ears in the model. The auger's transport capacity should be greater than the total number of fruit ears harvested by the harvester in the next second. Therefore, the auger's processing speed for the harvested objects is greater than the harvester's processing speed for the harvested objects as a constraint, that is: ; Determine the value range of the auger speed, namely: ; The minimum rotation speed of the auger is obtained based on the value range of the auger rotation speed.

[0067] In some embodiments of the present application, after obtaining the running speed of each processing component used to respectively perform the picking process, the method further includes: adjusting the output flow of the hydraulic pump through a control signal, controlling the hydraulic motor to drive each processing component to operate at the running speed; and obtaining the real-time speed of each working component of the cutting platform according to the current engine speed and feedback signal. Specifically, the cutting platform execution unit realizes the matching of the speed of the cutting platform working component with the travel speed and the agronomic parameters of corn through the control signal sent by the intelligent control unit, the engine outputs power to the hydraulic pump, the hydraulic pump converts mechanical energy into hydraulic energy, and adjusts the output flow according to the control signal to achieve precise control of the speed of the hydraulic motor, thereby achieving control of the cutting platform working component, and the electric proportional hydraulic variable pump can send a feedback signal back to the intelligent control unit, and the real-time speed of each working component of the cutting platform can be obtained according to the current engine speed and feedback signal.

[0068] Through the above implementation, the above-mentioned target detection model for corn stalks and corn ears, the corn agronomy calculation model based on binocular vision, and the calculation of the required rotation speed of each component of the cutting table are all integrated into the data processing module of the intelligent control unit to achieve an end-to-end rotation speed matching effect. In the actual working process of the corn machine, each time the binocular camera takes a picture, the image data is transmitted to the intelligent control unit through the image acquisition card. The intelligent control unit uses the target detection model to identify and mark the corn ears and corn stalks in the picture, and performs corn agronomy calculations. According to the calculation results, the corresponding minimum rotation speed matched by each component of the cutting table under the current working state and working conditions is determined, and the control signal is output to the actuator through the execution control module, thereby realizing accurate control of the working state of the actuator.

[0069] Based on the same inventive concept, the present application also provides a device for automatically matching the rotation speed of a header component. Figure 7 The schematic diagram of the structure of the automatic speed matching device of the header components according to the embodiment of the present application is shown schematically. Figure 7As shown, a device for automatically matching the rotation speed of a cutting table component includes: a parameter acquisition module, used to acquire agronomic parameters of the area to be picked in front of the cutting table; the agronomic parameters include: size information of the object to be picked, distribution information of the stem where the object to be picked is located, and at least one of the relative positions of the object to be picked on the stem where it is located; a processing speed module, which obtains the processing speed of the cutting table for the object to be picked according to the agronomic parameters and the moving speed of the cutting table; and a speed output module, which is used to obtain the operating speeds of various processing components of the cutting table used to respectively execute the picking process based on the processing speed of the cutting table for the object to be picked.

[0070] In some optional embodiments, the processing speed of the cutting platform for the objects to be picked is obtained according to the agronomic parameters and the travel speed of the cutting platform, including: obtaining the plant processing speed according to the travel speed and the stem distribution information of the objects to be picked in the agronomic parameters; and converting the plant processing speed into the processing speed of the cutting platform for the objects to be picked according to the number of objects to be picked per unit area.

[0071] In some optional embodiments, the processing components of the harvesting platform for executing the picking process include stem pulling rollers; the operating speeds of the various processing components of the harvesting platform for respectively executing the picking process are obtained based on the processing speeds, including: the stem pulling time is not greater than the driving time as a first constraint; the processing speed of the stem pulling roller to treat the picked objects is greater than the processing speed of the harvesting platform to treat the picked objects as a second constraint; the linear speed of the stem pulling roller that satisfies the first constraint and the second constraint is used as the minimum linear speed of the stem pulling roller.

[0072] In some optional implementations, the first constraint is expressed as: ; in, V 拉 is the linear speed of the stem pulling roller, s is the safety factor, H 结穗 is the ear height, H 摘穗 is the cutting height of the header, t 车 is the time required for the header to travel one plant spacing; the second constraint condition is expressed as: ; Among them, Q 拉 The processing speed of the stem roller to be picked is n is the number of rows at harvest time, and Q is the processing speed of the harvesting platform.

[0073] In some optional embodiments, the processing components of the harvesting platform for executing the picking process include a straw chain; based on the processing speed, the operating speeds of the various processing components of the harvesting platform for respectively executing the picking process are obtained, including: obtaining an expression for the processing speed of the straw chain for treating the picked objects, the parameters of the expression including the linear speed of the straw chain; determining the value range of the linear speed of the straw chain with the constraint that the processing speed of the straw chain for treating the picked objects is greater than the processing speed of the harvesting platform for treating the picked objects; and obtaining the minimum linear speed of the straw chain based on the value range of the linear speed of the straw chain.

[0074] In some optional embodiments, the expression of the processing speed of the reed chain on the objects to be picked includes: ; Among them, Q 拨 The speed at which the harvesting chain processes the harvested material. n is the number of rows at harvest, Ψ 1 is the filling coefficient. Considering that the ears in the picking path are not completely full, the filling coefficient is between 0.5 and 1. L 0 is the length of corn ear, V 拨 is the linear speed of the reed chain.

[0075] In some optional embodiments, the processing components of the harvesting platform for executing the picking process include an auger; based on the processing speed, the operating speeds of the various processing components of the harvesting platform for respectively executing the picking process are obtained, including: obtaining an expression for the processing speed of the auger for the picked objects, wherein the parameters of the expression include the auger speed; determining the range of the auger speed with the constraint that the processing speed of the auger for the picked objects is greater than the processing speed of the harvesting platform for the picked objects; and obtaining the minimum auger speed based on the range of the auger speed.

[0076] In some optional embodiments, the expression of the processing speed of the auger on the harvested material includes: ; Among them, Q 搅龙 To control the speed of the harvesting process, η is the actual efficiency value of the auger, P is the screw pitch of the auger, π is the circumference of a circle, d 1 is the outer diameter of the auger blade, d 2 is the inner diameter of the auger blade, V 1 is the volume of the material to be picked, n 搅龙 is the auger speed.

[0077] In some optional embodiments, agronomic parameters of the area to be harvested in front of the harvesting platform are obtained, including: obtaining a first image and a second image of the area to be harvested through a first camera and a second camera respectively; identifying a number of feature points of the object to be harvested and the stems on which they are located from the first image and the second image; determining depth information of the number of feature points according to position parameters and optical parameters of the first camera and the second camera; determining a target to be measured based on the agronomic parameters to be determined, and selecting feature points of the target to be measured from the number of feature points; and obtaining the value of the agronomic parameter to be determined based on the depth information of the feature points of the target to be measured and the pixel size information of the image of the target to be measured.

[0078] In some optional embodiments, several feature points of the object to be picked and the stem on which it is located are identified from the first image and the second image, including: using a trained real-time target detection algorithm to identify the object to be picked and the stem on which it is located from the first image and the second image respectively; using a scale-invariant feature transformation algorithm to identify a feature point set from the object to be picked and the stem on which it is located, to obtain a first feature point set corresponding to the first image and a second feature point set corresponding to the second image; matching the first feature point set with the second feature point set to obtain several feature points of the object to be picked and the stem on which it is located.

[0079] In some optional embodiments, the real-time target detection algorithm is obtained by improving the YOLO algorithm as follows: introducing a channel-priority convolutional attention module into the backbone layer of the YOLO algorithm; and introducing a multi-scale convolutional module into the connection layer of the YOLO algorithm.

[0080] In some optional embodiments, the depth information of the plurality of feature points is determined according to the position parameters and optical parameters of the first camera and the second camera, including: for each of the plurality of feature points, at least one set of similar triangles is constructed by selecting a plurality of points from among the feature point, the optical center of the first camera, the optical center of the second camera, the imaging point of the feature point in the first camera, and the imaging point of the feature point in the second camera; based on the similar triangles, a distance between the optical centers, the focal length of the first camera, the focal length of the second camera, and a plurality of line lengths from the distance between the imaging point and the end of the imaging surface is calculated to obtain a distance between the feature point and a line connecting the optical centers of the cameras or a distance between the feature point and the imaging surface; and the distance between the feature point and the optical center of the camera or the distance between the feature point and the imaging surface is used as the depth information of the feature point.

[0081] In some optional embodiments, the target to be measured is determined based on the agronomic parameters to be determined, and the feature points of the target to be measured are selected from the several feature points, including: if the agronomic parameter to be determined is the size information of the object to be picked, the target to be measured is the object to be picked, and the feature points of the target to be measured include the left feature point, the right feature point, the top feature point and the bottom feature point of the object to be picked; if the agronomic parameter to be determined is the distribution information of the stem where the object to be picked is located, the target to be measured is the stem where the object to be picked is located, and the feature points of the target to be measured include the bottom feature point of the stem where the object to be picked is located; if the agronomic parameter to be determined is the relative position of the object to be picked on the stem where it is located, the target to be measured is the object to be picked and the stem where it is located, and the feature points of the target to be measured include the bottom feature point of the picked object and the bottom feature point of the stem where the object to be picked is located.

[0082] In some optional embodiments, the value of the agronomic parameter to be determined is obtained based on the depth information of the feature points of the target to be measured and the pixel size information of the image of the target to be measured, including: obtaining the pixel distance between feature points based on the pixel size information of the line segment formed by the feature points of the target to be measured in the image; obtaining the projection height of the line segment formed by the feature points of the target to be measured based on the pixel distance and the depth information, and obtaining the value of the agronomic parameter to be determined using the projection height.

[0083] In some optional embodiments, when the depth information of different feature points of the target to be measured is inconsistent, the device also includes a projection correction module, which is used to: calculate the depth difference between different feature points in the same line segment, correct the projection height based on the depth difference, and obtain the value of the agronomic parameter to be determined by the corrected projection height.

[0084] In some optional embodiments, after obtaining the operating speeds of the various processing components used to respectively perform the picking process, the device also includes a control monitoring module, which is used to: adjust the output flow of the hydraulic pump through a control signal, control the hydraulic motor to drive each processing component to operate at the operating speed; and obtain the real-time speed of each working component of the cutting platform according to the current engine speed and feedback signal.

[0085] The specific definition of each functional module in the above-mentioned automatic matching device for the rotation speed of the cutting table components can be found in the above-mentioned definition of the automatic matching method for the rotation speed of the cutting table components, which will not be repeated here. Each module in the above-mentioned system can be implemented in whole or in part by software, hardware and a combination thereof. The above-mentioned modules can be embedded in or independent of the processor in the electronic device in the form of hardware, or can be stored in the memory in the electronic device in the form of software, so that the processor can call and execute the operations corresponding to the above modules. It also realizes the advantages of automatic control based on self-adaptation of agronomic parameters.

[0086] In some embodiments of the present application, an electronic device is also provided, comprising: at least one processor; a memory connected to the at least one processor; wherein the memory stores instructions that can be executed by the at least one processor, and the at least one processor executes the above-mentioned method for automatically matching the rotation speed of the header component. The internal structure diagram thereof can be as follows Figure 8 shown. Figure 8 The internal structure diagram of an electronic device according to an embodiment of the present application is schematically shown. The electronic device includes a processor A01, a network interface A02, a memory (not shown in the figure) and a database (not shown in the figure) connected via a system bus. Among them, the processor A01 of the electronic device is used to provide computing and control capabilities. The memory of the electronic device includes an internal memory A03 and a non-volatile storage medium A04. The non-volatile storage medium A04 stores an operating system B01, a computer program B02 and a database (not shown in the figure). The internal memory A03 provides an environment for the operation of the operating system B01 and the computer program B02 in the non-volatile storage medium A04. The network interface A02 of the electronic device is used to communicate with an external terminal through a network connection. When the computer program B02 is executed by the processor A01, a method for automatically matching the rotation speed of a header component is implemented.

[0087] Those skilled in the art will understand that Figure 8 The structure shown in the figure is merely a block diagram of a partial structure related to the scheme of the present application, and does not constitute a limitation on the electronic device to which the scheme of the present application is applied. The specific electronic device may include more or fewer components than shown in the figure, or combine certain components, or have a different arrangement of components.

[0088] In one embodiment provided in the present application, a harvesting machine is also provided. Fig. 9 The schematic diagram of the system structure of the harvesting machine according to the embodiment of the present application is shown schematically. Fig. 9 As shown, the harvesting machine includes: a visual recognition unit, a sensor unit, an intelligent control unit, a harvesting table execution unit, a human-machine interaction unit and a connecting harness.

[0089] Among them, the visual recognition unit includes a camera and a deep learning target detection model. The camera uses two area array cameras to form a binocular camera to take pictures. The target detection model is trained so that it can accurately identify corn ears and corn stalks, and is embedded in the intelligent control unit.

[0090] The sensor unit includes a speed sensor and an infrared sensor, which are respectively installed on the front axle and the cutting table of the corn harvester, and are used to monitor the driving speed and the height of the cutting table of the corn harvester in real time when the corn harvester is operating.

[0091] The cutter table execution unit is composed of an execution electric proportional hydraulic variable pump, an execution motor, and various working parts of the cutter table. The control signal sent by the intelligent control unit can achieve the matching of the speed of the cutter table working parts with the travel speed and corn agronomic parameters. The engine outputs power to the hydraulic pump, the hydraulic pump converts mechanical energy into hydraulic energy, and adjusts the output flow or the direction of oil inlet and outlet according to the control signal to achieve precise control of the speed of the hydraulic motor, thereby achieving control of the cutter table working parts. The electric proportional hydraulic variable pump can send feedback signals back to the intelligent control unit, and the real-time speed of each working part of the cutter table can be obtained according to the current engine speed and feedback signals; the control panel can display the monitored travel speed and the speed of each part of the cutter table in real time.

[0092] The core processor of the intelligent control unit is PLC, and its main function is to execute the aforementioned method for automatically matching the rotation speed of the cutting table components. It may include a data acquisition module, a data processing module and an execution control module. The data acquisition module can receive pictures taken by the camera, and can also read information such as the travel speed and the height of the cutting table from the sensor; the data processing module can recognize the taken pictures, and can use the built-in binocular vision algorithm to perform calculations and analysis on various agronomical calculations of corn, and then calculate the required rotation speed of each component of the cutting table, and send out a control signal; the execution control module outputs the control signal to the actuator, and then matches and adjusts the speed of each component of the cutting table.

[0093] The connection harness is a type of harness used to connect sensor units, intelligent control units, execution units, and human-machine interaction units, including communication lines, power lines, and signal lines. Communication lines are used between the intelligent control unit and the human-machine interaction unit and execution unit, and signal lines are used between the control unit and the sensor unit. All hardware needs to be directly / indirectly connected to the power line from the battery.

[0094] In general, the binocular camera is used to collect picture information of the working scene in real time, and the corn ears and corn stalks in the current working scene are detected based on the deep learning target detection algorithm. The current corn plant spacing, ear height, corn ear length and diameter are analyzed based on the binocular vision technology. The speed sensor and infrared sensor are used to monitor the current travel speed and header height. The corn processing capacity of the header per second is calculated based on the travel speed and plant spacing. The required speed of each component of the header is inferred based on the processing capacity and the number of harvested rows, corn plant spacing, ear height, corn ear length and diameter, and it is adjusted to achieve automatic matching of the speeds of each component of the header.

[0095] In one embodiment provided in the present application, a machine-readable storage medium is provided, on which instructions are stored. When the instructions are executed by a processor, the processor is configured to execute the aforementioned method for automatically matching the rotation speed of the header components.

[0096] In one embodiment provided in the present application, a computer program product is provided, including a computer program, which implements the aforementioned method for automatically matching the rotation speed of the header components when executed by a processor.

[0097] Those skilled in the art will appreciate that the embodiments of the present application may be provided as methods, systems, or computer program products. Therefore, the present application may adopt the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Moreover, the present application may adopt the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program codes.

[0098] The present application is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of the processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0099] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0100] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 A step that specifies a function in one or more boxes.

[0101] In a typical configuration, a computing device includes one or more processors (CPU), input / output interfaces, network interfaces, and memory.

[0102] The memory may include non-permanent memory in a computer-readable medium, random access memory (RAM) and / or non-volatile memory in the form of read-only memory (ROM) or flash RAM. The memory is an example of a computer-readable medium.

[0103] Computer readable media include permanent and non-permanent, removable and non-removable media that can be implemented by any method or technology to store information. Information can be computer readable instructions, data structures, program modules or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technology, compact disk read-only memory (CD-ROM), digital versatile disk (DVD) or other optical storage, magnetic cassettes, magnetic tape magnetic disk storage or other magnetic storage devices or any other non-transmission media that can be used to store information that can be accessed by a computing device. As defined herein, computer readable media does not include temporary computer readable media (transitory media), such as modulated data signals and carrier waves.

[0104] It should also be noted that the terms "include", "comprises" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, commodity or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, commodity or device. In the absence of more restrictions, the elements defined by the sentence "comprises a ..." do not exclude the existence of other identical elements in the process, method, commodity or device including the elements.

[0105] The above are only embodiments of the present application and are not intended to limit the present application. For those skilled in the art, the present application may have various changes and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application should be included within the scope of the claims of the present application.

Claims

1. A method for automatically matching the rotation speed of a header component, characterized in that: The method includes: Acquire agronomic parameters of the area to be picked in front of the header; the agronomic parameters include: at least one of: size information of the object to be picked, distribution information of the stem where the object to be picked is located, and relative position of the object to be picked on the stem where the object is located; According to the agronomic parameters and the travel speed of the harvesting platform, obtaining the processing speed of the harvesting platform for the objects to be harvested; Based on the processing speed of the harvesting platform on the objects to be harvested, the operating speeds of the various processing components of the harvesting platform for respectively executing the harvesting process are obtained.

2. The method for automatically matching the rotation speed of header components according to claim 1, characterized in that: According to the agronomic parameters and the travel speed of the header, the processing speed of the header for the objects to be harvested is obtained, including: Obtaining a plant processing speed according to the travel speed and the stem distribution information of the objects to be picked in the agronomic parameters; The plant processing speed is converted into a processing speed of the harvesting platform for the objects to be harvested according to the number of objects to be harvested per unit area.

3. The method for automatically matching the rotation speed of header components according to claim 1, characterized in that: The processing components of the harvesting platform for performing the picking process include a stem pulling roller; and the operating speeds of the various processing components of the harvesting platform for performing the picking process respectively are obtained based on the processing speed, including: The first constraint condition is that the stem pulling time should not be greater than the driving time; The second constraint condition is that the processing speed of the stem pulling roller for picking is greater than the processing speed of the cutting table for picking; The linear speed of the stem-pulling roller that satisfies the first constraint condition and the second constraint condition is taken as the minimum linear speed of the stem-pulling roller.

4. The method for automatically matching the rotation speed of a header component according to claim 3, characterized in that: The first constraint is expressed as: ； in, V 拉 is the linear speed of the stem pulling roller, s is the safety factor, H 结穗 is the ear height, H 摘穗 is the cutting height of the header, t 车 The time required for the header to travel one plant spacing; The second constraint is expressed as: ； Among them, Q 拉 The processing speed of the stem roller to be picked is n is the number of rows at harvest time, and Q is the processing speed of the harvesting platform.

5. The method for automatically matching the rotation speed of header components according to claim 1, characterized in that: The processing components of the harvesting platform for performing the harvesting process include a reed chain; and the operating speeds of the various processing components of the harvesting platform for performing the harvesting process respectively are obtained based on the processing speed, including: Obtaining an expression for the processing speed of the reed-pulling chain on the objects to be picked, wherein the parameters of the expression include the linear speed of the reed-pulling chain; Taking the processing speed of the reed chain for the picked objects as a constraint condition being greater than the processing speed of the header for the picked objects, determining the value range of the linear speed of the reed chain; The minimum linear speed of the reel chain is obtained based on the value range of the linear speed of the reel chain.

6. The method for automatically matching the rotation speed of a header component according to claim 5, characterized in that: The expression of the processing speed of the reed chain for the picked objects includes: ； Among them, Q 拨 The speed at which the harvesting chain processes the harvested material. n is the number of rows at harvest, Ψ 1 is the filling coefficient. Considering that the ears in the picking path are not completely full, the filling coefficient is between 0.5 and 1. L 0 is the length of corn ear, V 拨 is the linear speed of the reed chain.

7. The method for automatically matching the rotation speed of header components according to claim 1, characterized in that: The processing components of the harvesting platform for performing the harvesting process include a auger; and the operation speeds of the various processing components of the harvesting platform for respectively performing the harvesting process are obtained based on the processing speed, including: Obtaining an expression for the processing speed of the auger on the objects to be picked, wherein the parameters of the expression include the auger rotation speed; Taking the processing speed of the auger for the harvested objects as a constraint condition, which is greater than the processing speed of the header for the harvested objects, the value range of the auger rotation speed is determined; The minimum rotation speed of the auger is obtained based on the value range of the auger rotation speed.

8. The method for automatically matching the rotation speed of header components according to claim 7, characterized in that: The expression for the speed at which the auger is handling the harvested material includes: ； Among them, Q 搅龙 To control the speed of the harvesting process, η is the actual efficiency value of the auger, P is the screw pitch of the auger, π is the circumference of a circle, d 1 is the outer diameter of the auger blade, d 2 is the inner diameter of the auger blade, V 1 is the volume of the material to be picked, n 搅龙 is the auger speed.

9. The method for automatically matching the rotation speed of header components according to claim 1, characterized in that: Obtain agronomic parameters of the area to be harvested in front of the header, including: Acquire a first image and a second image of the area to be picked by using a first camera and a second camera respectively; Identifying several feature points of the object to be picked and the stem on which it is located from the first image and the second image; Determine depth information of the plurality of feature points according to position parameters and optical parameters of the first camera and the second camera; Determining a target to be measured based on the agronomic parameter to be determined, and selecting a feature point of the target to be measured from the plurality of feature points; The value of the agronomic parameter to be determined is obtained based on the depth information of the feature points of the target to be measured and the pixel size information of the image of the target to be measured.

10. The method for automatically matching the rotation speed of header components according to claim 9, characterized in that: Identifying several feature points of the object to be picked and the stem on which it is located from the first image and the second image, including: Using a trained real-time target detection algorithm to identify the object to be picked and the stem on which it is located from the first image and the second image respectively; A scale-invariant feature conversion algorithm is used to identify a feature point set from the object to be picked and the stem on which it is located, so as to obtain a first feature point set corresponding to the first image and a second feature point set corresponding to the second image; The first feature point set and the second feature point set are matched to obtain a plurality of feature points of the object to be picked and the stem on which it is located.

11. The method for automatically matching the rotation speed of header components according to claim 9, characterized in that: The real-time target detection algorithm is obtained by improving the YOLO algorithm as follows: Introducing a channel-priority convolutional attention module into the backbone layer of the YOLO algorithm; and A multi-scale convolution module is introduced in the connection layer of the YOLO algorithm.

12. The method for automatically matching the rotation speed of a header component according to claim 9, characterized in that: Determining depth information of the plurality of feature points according to position parameters and optical parameters of the first camera and the second camera includes: For each of the plurality of feature points, construct at least one set of similar triangles by selecting a plurality of points from the feature point, the optical center of the first camera, the optical center of the second camera, the imaging point of the feature point in the first camera, and the imaging point of the feature point in the second camera; Based on the similar triangles, the distance between the feature point and the line connecting the optical centers of the cameras or the distance between the feature point and the imaging surface is calculated by using the distance between the optical centers, the focal length of the first camera, the focal length of the second camera, and the distance between the imaging point and the end of the imaging surface; The distance between the feature point and the line connecting the optical center of the camera or the distance between the feature point and the imaging plane is used as the depth information of the feature point.

13. The method for automatically matching the rotation speed of a header component according to claim 9, characterized in that: Determining a target to be measured based on the agronomic parameter to be determined, and selecting a feature point of the target to be measured from the plurality of feature points, including: If the agronomic parameter to be determined is the size information of the object to be picked, the object to be measured is the object to be picked, and the characteristic points of the object to be measured include the left characteristic point, the right characteristic point, the top characteristic point and the bottom characteristic point of the object to be picked; If the agronomic parameter to be determined is the distribution information of the stem where the object to be picked is located, the target to be measured is the stem where the object to be picked is located, and the characteristic point of the target to be measured includes the bottom characteristic point of the stem where the object to be picked is located; If the agronomic parameter to be determined is the relative position of the object to be picked on its stem, the target to be measured is the object to be picked and its stem, and the feature points of the target to be measured include the bottom feature point of the object to be picked and the bottom feature point of the stem where the object to be picked is located.

14. The method for automatically matching the rotation speed of a header component according to claim 9, characterized in that: Obtaining the value of the agronomic parameter to be determined based on the depth information of the feature points of the target to be measured and the pixel size information of the image of the target to be measured, including: Obtaining the pixel distance between the feature points based on the pixel size information of the line segments formed by the feature points of the target to be measured in the image; The projection height of the line segment formed by the characteristic points of the target to be measured is obtained based on the pixel distance and the depth information, and the value of the agronomic parameter to be determined is obtained by the projection height.

15. The method for automatically matching the rotation speed of a header component according to claim 14, characterized in that: When the depth information of different feature points of the target to be measured is inconsistent, the method further includes: The depth difference between different feature points in the same line segment is calculated, the projection height is corrected based on the depth difference, and the value of the agronomic parameter to be determined is obtained by using the corrected projection height.

16. The method for automatically matching the rotation speed of header components according to claim 1, characterized in that: After obtaining the operating speeds of the various processing components for respectively performing the picking process, the method further includes: Adjusting the output flow of the hydraulic pump through the control signal, controlling the hydraulic motor to drive each processing component to operate at the operating speed; and The real-time speed of each working part of the header is obtained according to the current engine speed and feedback signal.

17. A device for automatically matching the rotation speed of a header component, characterized in that: The device includes: A parameter acquisition module, used to acquire agronomic parameters of the area to be picked in front of the header; the agronomic parameters include: at least one of the size information of the object to be picked, the distribution information of the stem where the object to be picked is located, and the relative position of the object to be picked on the stem where it is located; a processing speed module, which obtains a processing speed of the harvesting table on the objects to be harvested according to the agronomic parameters and the travel speed of the harvesting table; and The speed output module is used to obtain the operating speed of each processing component of the harvesting platform for respectively executing the harvesting process based on the processing speed of the harvesting platform on the objects to be harvested.

18. An electronic device, characterized in that: include: at least one processor; a memory connected to the at least one processor; The memory stores instructions that can be executed by the at least one processor, and the at least one processor implements the steps of the method for automatic matching of the rotation speed of the header components as described in any one of claims 1 to 16 by executing the instructions stored in the memory.

19. A computer-readable storage medium having a computer program / instruction stored thereon, characterized in that: When the computer program / instructions are executed by a processor, the steps of the method for automatically matching the rotation speed of the header components described in any one of claims 1 to 16 are implemented.

20. A harvesting machine, characterized in that: The PLC in the harvesting machine is configured to execute the method for automatic matching of the rotation speed of the header components as described in any one of claims 1 to 16.

Images