A rice harvesting property detection and operation regulation system and method and a harvester

Abstract

The application provides a rice harvesting attribute detection and operation regulation system and method and a harvester. An image acquisition device is used to acquire the rice harvesting attribute of a left area to be harvested during the operation of the harvester, and the attribute information is stored after being matched with the position information. When the harvester reaches a new position, the rice harvesting attribute is recalled according to the current position, the forward speed is controlled by combining a forward speed control model, and accurate rice harvesting is realized. The control method of the rice harvesting attribute detection and operation regulation system adopts a semantic segmentation, skeleton line extraction and image depth method to detect the harvesting attribute of rice, changes the forward speed according to different harvesting attributes and forward speed control models, and realizes accurate control of the forward speed. The harvester adopts the control method of the rice harvesting attribute detection and operation regulation system to control the rice harvesting attribute detection and operation regulation system, and realizes accurate rice harvesting.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of rice combine harvesters, in particular to a rice harvesting attribute detection and operation regulation system and method and a harvester. BACKGROUND

[0002] Unmanned driving of harvesting machinery can improve work efficiency and reduce work intensity, but current unmanned driving of harvesting machinery mainly studies extraction and tracking of navigation paths, and the growth environment of rice in farmland is complex and changeable, and the growth of rice is affected by many factors such as light, moisture, soil, pests and diseases, which leads to differences in plant height and density of rice in the same farmland. When the combine harvester is in a stable forward speed and cutting width state, due to the different plant height and density of the rice in the area to be harvested, the feeding amount of the harvester will fluctuate. Excessive feeding amount increases the load of the threshing cylinder and other working parts, at which time the harvester is prone to jamming, and insufficient feeding amount will cause insufficient load of the threshing cylinder, resulting in insufficient threshing of rice grains, increasing loss and reducing work efficiency. Therefore, the unmanned operation of the combine harvester not only needs to realize autonomous walking and harvesting, but also needs to adjust the forward speed of the harvester according to the crop density, plant height and other crop attribute information to ensure the stability of the feeding amount.

[0003] Patent CN115777330A proposes a grain combine harvester feeding amount detection method, which detects the feeding amount by installing sensors on the outlet bottom plate of the chain rake conveyor and on the driving wheel of the chain rake conveyor, but the detected data occurs after the crop is fed into the harvester, cannot perceive the feeding amount information in advance, and the measurement accuracy is not high, the sensor is easily affected by vehicle body vibration, and it is difficult to provide effective data for real-time regulation; Patent CN110235600A proposes a combine harvester feeding amount stable control system based on mature crop attribute information real-time detection, which controls the operation speed by detecting the attribute information of the crop to be harvested in front of the combine harvester during operation, and the point cloud and image processing calculation amount is large, the regulation is carried out at the same time of detection, and the regulation system is difficult to respond quickly.

[0004] In order to overcome the shortcomings of low precision, large calculation amount and lagging of the traditional harvesting attribute detection method and operation regulation system, the present application provides a rice harvesting attribute detection and operation regulation system and method and a harvester. SUMMARY

[0005] In view of the above technical problems, one of the purposes of one embodiment of the present application is to provide a rice harvesting attribute detection and operation control system, which acquires the rice harvesting attribute of the area to be harvested during the operation of the harvester by using an image acquisition device, stores the attribute information and the position information after correspondence, and retrieves the rice harvesting attribute according to the current position during operation, and controls the forward speed by combining the forward speed control model to realize accurate rice harvesting.

[0006] The present application provides a rice harvesting attribute detection and operation control system, which acquires the rice harvesting attribute of the left area to be harvested during the operation of the harvester by using an image acquisition device, and stores the attribute information and the position information after correspondence, and retrieves the rice harvesting attribute according to the current position when the harvester reaches a new position, and controls the forward speed by combining the forward speed control model to realize accurate rice harvesting.

[0007] The present application also provides a control method of a rice harvesting attribute detection and operation control system, which detects the harvesting attribute of rice by using semantic segmentation, skeleton line extraction, and image depth method, changes the forward speed according to different harvesting attributes and forward speed control model, and realizes accurate control of the forward speed.

[0008] The present application also provides a harvester, which controls the rice harvesting attribute detection and operation control system by using the control method of the rice harvesting attribute detection and operation control system, and realizes accurate harvesting of rice.

[0009] Note that the description of these purposes does not hinder the existence of other purposes. One embodiment of the present application does not need to realize all the above purposes. The purposes other than the above purposes can be extracted from the description, drawings, and claims.

[0010] The present application realizes the above technical purposes by the following technical means:

[0011] A rice harvesting attribute detection and operation control system, comprising an image acquisition device, a positioning device, an information processing system, and a forward speed control module;

[0012] The image acquisition device is used to acquire the image of rice in the area to be harvested on one side of the harvester, and the image acquisition device is connected with the information processing system and transmits the acquired image of rice to the information processing system;

[0013] The positioning device is used to acquire the position and heading information of the harvester, and the positioning device is connected with the information processing system and transmits the position and heading information to the information processing system;

[0014] The information processing system is used for processing the rice image collected by the image collection device, extracting the rice harvesting attribute information, and processing the position and heading information of the harvester obtained by the positioning device, obtaining the position information of the collected rice in the area to be harvested through coordinate conversion, and storing the rice harvesting attribute information and the position information one by one; when the harvester reaches a new position, the rice harvesting attribute information of the current harvesting position is called, the forward speed is calculated, and the forward speed control module is transmitted.

[0015] The forward speed control module is connected with the information processing system, and controls the forward speed of the harvester according to the forward speed control instruction issued by the information processing system.

[0016] In the above scheme, the image collection device is a binocular camera, which is vertically downward;

[0017] The positioning device is an RTK-GNSS system;

[0018] The information processing system is an embedded development board or an industrial computer.

[0019] In the above scheme, the forward speed of the harvester issued by the information processing system is calculated according to the rice harvesting attribute matching the harvester forward speed control model, which is used to control the forward speed of the harvester; the forward speed control model is:

[0020] V = λ1 × D + λ2 × H + C2

[0021] Wherein, V is the forward speed, D is the density of the rice in the ROI area, λ1 is the density proportion coefficient, H is the plant height, λ2 is the plant height proportion coefficient, and C2 is a constant.

[0022] In the above scheme, the rice harvesting attribute refers to the density and plant height of the rice.

[0023] Further, the density of the rice is calculated by a rice density calculation model; the rice density calculation model is:

[0024] D = α × p + β × n + C1

[0025] Wherein, D is the rice density, p is the pixel area of the rice ear mask, n is the number of skeleton line nodes, α is the coefficient of p, β is the coefficient of n, and C1 is a constant.

[0026] A control method according to the rice harvesting attribute detection and operation control system, comprising the following steps:

[0027] Step S1, image collection: collecting the rice image in the area to be harvested on the left side of the harvester by the image collection device, obtaining the color RGB image and the depth image, and transmitting to the information processing system;

[0028] Step S2, position acquisition: measure the latitude and longitude coordinates of the current harvester position and the heading through the positioning device;

[0029] Step S3, information processing: process the rice image obtained in step S1 through the information processing system to obtain the plant height and density information of the rice, and calculate the forward speed of the harvester through the control model; process the harvester position and heading information, and obtain the position information of the left side of the harvested area through the latitude and longitude coordinate conversion, and store the rice harvesting attribute information and the position information one by one; when the harvester reaches a new position, the rice harvesting attribute information of the current harvesting position is called, the forward speed is calculated according to the forward speed control model, and the forward speed is transmitted to the forward speed control module;

[0030] Step S4, speed control: the forward speed control module controls the forward speed of the harvester according to the forward speed sent in step S3.

[0031] In the above scheme, the latitude and longitude coordinate conversion formula is:

[0032]

[0033]

[0034] where (x1, y1) is the latitude and longitude coordinates of the current position of the harvester, (x2, y2) is the latitude and longitude coordinates of the rice position photographed by the camera of the harvester, R is the radius of the earth, L is the distance between the two positions, and θ is the heading angle of the harvester, is the included angle between the line connecting the two positions and the heading of the harvester.

[0035] In the above scheme, the rice harvesting attribute refers to the density and plant height of the rice; the image acquisition device is a binocular camera, and the color RGB image and the depth image collected by the binocular camera are used as the input of the image processing module of the information processing system, and are used for obtaining the density and plant height information of the rice, respectively.

[0036] In the above scheme, the rice density acquisition method comprises the following steps:

[0037] According to the RGB image of the rice in the harvesting machine side of the harvested area collected by the image acquisition device, the ROI region is extracted with the harvesting machine cutting width as the scale;

[0038] The image in the ROI region is cut and subjected to histogram equalization processing;

[0039] Then the processed image is input into the final semantic segmentation network model to obtain a mask containing only the rice panicle;

[0040] The pixel area of the rice panicle mask is calculated;

[0041] extract the spike skeleton line using a skeleton line extraction algorithm, and calculate the number of skeleton line nodes;

[0042] The pixel area of the rice spike and the number of skeleton line nodes are taken as inputs, and the rice density is taken as output, and the rice density is calculated according to the rice density calculation model.

[0043] In the above scheme, the image in the ROI region is intercepted and histogram equalization processing is performed, and a rice spike image training set, a verification set and a test set are established using the processed image, and the spike in the training set and the test set is labeled; the labeled training set, the verification set and the unlabeled test set are all data augmented; the augmented training set is trained to obtain a semantic segmentation network model capable of extracting the spike; the augmented verification set is input into the trained semantic segmentation network model for verification, and the model parameters are optimized; the test set is input into the optimized semantic segmentation network model for testing to obtain the final semantic segmentation network model.

[0044] In the above scheme, the spike skeleton line is extracted using a skeleton line extraction algorithm, the skeleton line is dilated through morphological processing to expand the width of the skeleton line, and a pixel area is formed at the intersection of the skeleton line; the dilated skeleton line is compared with the original skeleton line, and the nodes are marked; noise points and misdetected nodes are filtered based on the domain features of the nodes, and the number of skeleton line nodes is calculated.

[0045] In the above scheme, the pixel area of the rice spike and the number of skeleton line nodes are taken as inputs, and the rice density is taken as output, and the rice spike pixel area and the number of skeleton line nodes obtained by processing the collected image data and the actual rice density are linearly plane fitted to obtain a rice density calculation model.

[0046] In the above scheme, the rice plant height acquisition method comprises the following steps:

[0047] By extracting the depth information of the rice surface in the ROI region to the camera, the average depth value h1 is obtained after filtering out the abnormal values of the depth information through an abnormal value filtering method, and the difference between h1 and the camera installation height h2 is obtained to obtain the rice plant height H = h2-h1.

[0048] In the above scheme, the forward speed control model is obtained by linearly plane fitting the rice density and plant height data obtained by processing the collected image and the actual forward speed in the working process.

[0049] A harvester comprising the rice harvesting attribute detection and operation regulation system, and the rice harvesting attribute detection and operation regulation system detects and controls according to the control method of the rice harvesting attribute detection and operation regulation system.

[0050] Compared with the prior art, the present application has the following beneficial effects:

[0051] According to one mode of the present application, a rice harvesting attribute detection and operation control system acquires the rice harvesting attribute of the left area to be harvested during the operation of the harvester by using an image acquisition device, stores the attribute information and the position information after correspondence, and controls the forward speed according to the current position to realize accurate rice harvesting.

[0052] According to one mode of the present application, a control method of a rice harvesting attribute detection and operation control system detects the harvesting attribute of rice by using semantic segmentation, skeleton line extraction, and image depth method, changes the forward speed according to different harvesting attributes and forward speed control model, and realizes accurate control of the forward speed.

[0053] According to one mode of the present application, a control method of a rice harvesting attribute detection and operation control system controls the rice harvesting attribute detection and operation control system of the harvester, and realizes accurate harvesting of rice.

[0054] Note that the description of these effects does not preclude the existence of other effects. One mode of the present application does not necessarily have all the above-mentioned effects. Effects other than the above-mentioned effects can be clearly seen and extracted from the description, drawings, claims, etc. BRIEF DESCRIPTION OF DRAWINGS

[0055] Figure 1 is a structural schematic diagram of one embodiment of the present application;

[0056] Figure 2 is a flowchart of a rice harvesting attribute detection and operation control system according to one embodiment of the present application;

[0057] Figure 3 is a flowchart of rice density detection according to one embodiment of the present application;

[0058] Figure 4 is a schematic diagram of rice density detection according to one embodiment of the present application.

[0059] In the figure: 1, image acquisition device; 2, positioning device; 3, information processing system; 4, forward speed control module. DETAILED DESCRIPTION

[0060] The embodiments of the present application will be described in detail below, and examples of the embodiments are shown in the accompanying drawings, wherein the same or similar reference numerals represent the same or similar elements or elements having the same or similar functions throughout. The embodiments described below by referring to the accompanying drawings are exemplary and are intended to explain the present application, and cannot be understood as a limitation of the present application.

[0061] In the description of the present application, it is to be understood by the terms "center", "longitudinal", "lateral", "length", "width", "thickness", "front", "back", "left", "right", "up", "down", "axial", "radial", "vertical", "horizontal", "inner", "outer" and the like indicate the orientation or positional relationship based on the orientation or positional relationship shown in the drawings, only for the convenience of describing the present application and simplifying the description, and do not indicate or imply that the devices or elements referred to must have a particular orientation, be constructed and operated in a particular orientation, and therefore cannot be understood as limiting the present application. In addition, the terms "first", "second" are only for the purpose of description, and cannot be understood as indicating or implying relative importance or implicitly indicating the number of technical features indicated. Therefore, the features defined as "first", "second" can be explicitly or implicitly included one or more of the features. In the description of the present application, the meaning of "a plurality of" is two or more, unless otherwise explicitly specified and limited.

[0062] In the present application, unless otherwise explicitly specified and limited, the terms "mounting", "connection", "connection", "fixing" and the like should be broadly understood, for example, it can be fixedly connected, or it can be detachably connected, or integrally connected; it can be mechanically connected, or it can be electrically connected; it can be directly connected, or it can be indirectly connected through an intermediate medium, it can be the communication inside two elements. For those skilled in the art, the specific meaning of the above terms in the present application can be understood according to the specific circumstances.

[0063] Embodiment 1

[0064] As Figure 1 shown, a preferred embodiment of the rice harvesting attribute detection and operation control system comprises an image acquisition device 1, a positioning device 2, an information processing system 3, and a forward speed control module 4.

[0065] The image acquisition device 1 is used to acquire the image of the rice in the left side of the harvester to be harvested, and the image acquisition device 1 is connected with the information processing system 3 and transmits the acquired rice image to the information processing system 3.

[0066] The positioning device 2 is used to acquire the position and heading information of the harvester, and the positioning device 2 is connected with the information processing system 3 and transmits the position and heading information to the information processing system 3.

[0067] The information processing system 3 is used for processing the rice image collected by the image collection device 1 to extract the rice harvesting attribute information; processing the position and heading information of the harvester obtained by the positioning device 2, obtaining the position information of the left side of the harvested area through coordinate conversion, and storing the rice harvesting attribute information and the position information one by one; when the harvester reaches a new position, the rice harvesting attribute information of the current harvesting position is called, the forward speed is calculated, and the forward speed control module 4 is transmitted.

[0068] The forward speed control module 4 is connected with the information processing system 3, and controls the forward speed of the harvester according to the forward speed control instruction issued by the information processing system.

[0069] Preferably, the image collection device 1 is a binocular camera, which is vertically downwardly photographed.

[0070] Preferably, the positioning device 2 is an RTK-GNSS system.

[0071] Preferably, the information processing system 3 is an embedded development board or an industrial computer.

[0072] The forward speed of the harvester issued by the information processing system 3 is calculated according to the rice harvesting attribute matching the harvester forward speed automatic control model, and is used for controlling the forward speed of the harvester. The forward speed control model is specifically:

[0073] V = λ1 × D + λ2 × H + C2

[0074] Wherein, V is the forward speed, D is the density of the rice in the ROI area, λ1 is the density proportion coefficient, H is the plant height, λ2 is the plant height proportion coefficient, and C2 is a constant.

[0075] The rice harvesting attribute refers to the density and plant height of the rice.

[0076] The density of the rice is calculated by the rice density calculation model; the rice density calculation model is specifically:

[0077] D = α × p + β × n + C1

[0078] Wherein, D is the density of the rice, p is the pixel area of the rice ear mask, n is the number of skeleton line nodes, α is the coefficient of p, β is the coefficient of n, and C1 is a constant.

[0079] The plant height of the rice H = h2-h1, that is, the depth information of the rice surface to the camera in the ROI area is extracted, the abnormal value of the depth information is filtered out by the abnormal value filtering method, and the average depth value h1 is taken, and the camera installation height h2 is subtracted, to obtain the plant height H = h2-h1 of the rice.

[0080] The harvester forward speed control module 4 is connected with the harvester power module, and controls the harvester forward speed.

[0081] As Figure 2 The control method of the rice harvesting attribute detection and operation regulation system comprises the following steps:

[0082] Step S1, image acquisition: the image acquisition device 1 is used to acquire the rice image in the left side area to be harvested of the harvester, to obtain the color RGB image and the depth image, and to transmit to the information processing system;

[0083] Step S2, position acquisition: the positioning device 2 is used to measure the latitude and longitude coordinates and the heading of the current position A of the harvester;

[0084] Step S3, information processing: the information processing system 3 is used to process the rice image obtained in step S1, to obtain the plant height and density information of the rice, and to calculate the forward speed of the harvester through the forward speed control model; the position A and the heading information of the harvester are processed, the position B information of the collected rice in the left side area to be harvested is obtained through coordinate conversion, and the rice harvesting attribute information and the position B information are stored in one-to-one correspondence; when the harvester header starts to harvest the rice in position B, the rice harvesting attribute information of the current harvesting position is retrieved, the forward speed is calculated according to the forward speed control model, and the forward speed is transmitted to the forward speed control module 4;

[0085] Step S4, speed control: the forward speed control module 4 controls the forward speed of the harvester according to the forward speed sent in step S3.

[0086] The rice harvesting attribute refers to the density and plant height of the rice; the image acquisition device is a binocular camera, and the color RGB image and the depth image collected by the binocular camera are used as the input of the image processing module of the information processing system, and are used for obtaining the density and plant height information of the rice, respectively;

[0087] The step S2 of position acquisition is specifically:

[0088] 1) the latitude and longitude information (x1, y1) and the heading θ of the current position of the harvester are obtained through RTK-GNSS;

[0089] 2) the latitude and longitude information (x2, y2) of the position of the collected rice image is obtained according to the coordinate conversion formula, and the coordinate conversion formula is:

[0090]

[0091]

[0092] Wherein, R is the radius of the earth, and L is the distance between point A and point B.

[0093] In this embodiment, R = 6.378 x 10 6 m, L = 1800 mm,

[0094] As shown in Figure 3 and 4 , the rice density acquisition method includes the following steps:

[0095] According to the RGB image of the rice in the harvesting machine side to be harvested area collected by the image acquisition device, the ROI region is extracted with the harvesting machine cutting width as the scale;

[0096] The image in the ROI region is intercepted and histogram equalization processing is performed;

[0097] The processed image is used to establish a rice ear image training set, a verification set and a test set, and the ears in the training set and the test set are labeled; the labeled training set, verification set and unlabeled test set are all data augmented; the augmented training set is trained to obtain a semantic segmentation network model capable of extracting the ear; the augmented verification set is input into the trained semantic segmentation network model for verification, and the model parameters are optimized; the test set is input into the optimized semantic segmentation network model for testing, and the final semantic segmentation network model is obtained;

[0098] Then the processed image is input into the final semantic segmentation network model to obtain a mask containing only the rice ear;

[0099] The pixel area of the rice ear mask is calculated;

[0100] The skeleton line extraction algorithm is used to extract the ear skeleton line;

[0101] The skeleton line is dilated by morphological processing to expand the width of the skeleton line, forming a pixel area at the intersection of the skeleton line; the dilated skeleton line is compared with the original skeleton line, and the nodes are marked;

[0102] Based on the domain features of the nodes, the noise points and the misdetected nodes are filtered out, and the number of skeleton line nodes is calculated;

[0103] The pixel area of the rice ear and the number of skeleton line nodes are taken as input, and the rice density is taken as output. The pixel area of the rice ear and the number of skeleton line nodes obtained by processing the collected image data and the actual rice density are linearly plane fitted to obtain a rice density calculation model;

[0104] According to the rice density calculation model, the rice density is calculated.

[0105] According to the present embodiment, preferably, the rice density acquisition method in step S3 specifically includes the following steps:

[0106] 1) According to the RGB image of the rice in the left area to be harvested of the harvester collected by the image acquisition device, the ROI region is extracted with the harvester cutting width as the scale (2000x1000);

[0107] 2) The image in the ROI region is intercepted and histogram equalization processing is performed:

[0108]

[0109] Wherein, g [0, 255] is the current image value; f(g) is the converted image pixel value; i is a non-negative integer; n(i) is the pixel number of the i-th gray level; m=2000 and n=1000 are the length and width of the image respectively.

[0110] 3) 800 training sets, 100 verification sets and 100 test sets of rice ear images are established using the processed image, and the ears in the training set and the test set are labeled; the labeled training set, verification set and unlabeled test set are all data augmented to 2400 training sets, 300 verification sets and 300 test sets; the augmented training set is input into the Unet basic network model for training to obtain a semantic segmentation network model capable of extracting the ear; the augmented verification set is input into the trained semantic segmentation network model for verification to optimize the model parameters; the test set is input into the optimized semantic segmentation network model for testing to obtain the final semantic segmentation network model;

[0111] 4) The processed image is input into the final semantic segmentation network model to obtain a red mask containing only the rice ear and calculate the pixel area of the rice ear mask;

[0112] 5) The skeleton line extraction algorithm is used to extract the ear skeleton line; the skeleton line is dilated by morphological processing to expand the width of the skeleton line, forming a pixel area at the intersection of the skeleton line; the dilated skeleton line is compared with the original skeleton line, and the nodes are marked; based on the domain features of the nodes, the noise points and the misdetected nodes are filtered out, and the number of skeleton line nodes is calculated;

[0113] 6) The pixel area of the rice ear and the number of skeleton line nodes are taken as the input, and the rice density is taken as the output, the pixel area of the rice ear and the number of skeleton line nodes obtained by processing the collected image data and the actual rice density are linearly plane fitted to obtain a rice density calculation model; the rice density is calculated according to the rice density calculation model, and the rice density calculation model is specifically:

[0114] D=α×p+β×n+C1

[0115] Wherein, D is the density of rice, alpha is the coefficient of p, beta is the coefficient of n, p is the pixel area of rice panicle mask, n is the number of skeleton line nodes, C1 is a constant, C1 is related to the pixel area, skeleton line node and corresponding rice density.

[0116] In the embodiment, alpha = 1.806 x 10 -4 , p = 37362, beta = 0.104, n = 322, C1 = 356.7, D = 397 panicles / m 2 ;

[0117] The method for obtaining the height of the rice plant in the step S3 comprises the following steps:

[0118] By extracting the depth information of the rice surface in the ROI region to the camera, the average depth value h1 is obtained after filtering the abnormal values of the depth information by the abnormal value filtering method, and the difference between h1 and the camera installation height h2 is obtained, so as to obtain the height H = h2-h1 of the rice plant.

[0119] According to the embodiment, the method for obtaining the height of the rice plant comprises the following specific steps:

[0120] By extracting the depth information of the rice surface in the ROI region to the camera, the average depth value h1 = 820 mm is obtained after filtering the abnormal values of the depth information by the abnormal value filtering method, and the difference between h1 and the camera installation height h2 = 2000 mm is obtained, so as to obtain the height H = h2-h1 = 1180 mm of the rice plant.

[0121] The forward speed of the harvester is calculated according to the forward speed control model of the harvester matched with the rice harvesting attribute at the current position, and is used for controlling the forward speed of the harvester, and the forward speed control model is obtained by linear plane fitting of the rice density and height data obtained by processing the collected image and the forward speed in the actual operation process, and specifically is:

[0122] V = lambda1 x D + lambda2 x H + C2

[0123] Wherein, V is the forward speed, D is the density of the rice in the ROI region, lambda1 is the density proportion coefficient, H is the height, lambda2 is the height proportion coefficient;

[0124] In the embodiment, lambda1 = 1.2 x 10 -2 , lambda2 = 2.7 x 10 -3 , C2 = 6.43, V = 1.52 m / s;

[0125] According to the embodiment, preferably, the image acquisition device is used to obtain the rice harvesting attribute of the left area to be harvested in the operation process of the harvester, and the forward speed control model of the rice harvesting attribute and the forward speed is established to control the forward speed, so as to realize the accurate harvesting of the rice.

[0126] The application adopts semantic segmentation, skeleton line extraction, and image depth method to calculate and store the harvesting properties of the rice in the left area to be harvested of the harvester, matches the detected harvesting properties according to the current position of the harvester, and controls the advancing speed in combination with the advancing speed control model to realize the accurate harvesting of the rice.

[0127] Embodiment 2

[0128] A harvester comprising the operation regulation system of the rice harvesting property detection method, thus having the beneficial effects described in Embodiment 1, which will not be repeated here.

[0129] It should be understood that although the present specification is described in terms of various embodiments, not every embodiment contains only one independent technical solution, and the description of the specification is only for the sake of clarity. Those skilled in the art should consider the specification as a whole, and the technical solutions in each embodiment can also be combined appropriately to form other embodiments that can be understood by those skilled in the art.

[0130] The series of detailed descriptions listed above are only specific descriptions of the feasible embodiments of the present application, and are not intended to limit the protection scope of the present application. Any equivalent embodiments or changes made without departing from the spirit of the present application should be included in the protection scope of the present application.

Claims

1. A control method of a rice harvesting property detection and operation regulation system, characterized by, The system comprises an image acquisition device (1), a positioning device (2), an information processing system (3), and an advance speed control module (4). The image acquisition device (1) is used to acquire the image of the rice in the area to be harvested on the side of the harvester, and is connected with the information processing system (3) and transmits the acquired image of the rice to the information processing system (3). The positioning device (2) is used to acquire the position and heading information of the harvester, and is connected with the information processing system (3) and transmits the position and heading information to the information processing system (3). The information processing system (3) is used to process the image of the rice acquired by the image acquisition device (1), extract the harvesting attribute information of the rice, and process the position and heading information of the harvester acquired by the positioning device, obtain the position information of the rice in the area to be harvested through coordinate conversion, and store the harvesting attribute information of the rice and the position information in one-to-one correspondence. When the harvester reaches a new position, the harvesting attribute information of the current harvesting position is retrieved, the advance speed is calculated, and the advance speed control module (4) is transmitted. The advance speed control module (4) is connected with the information processing system (3), and controls the advance speed of the harvester according to the advance speed control instruction issued by the information processing system (3). The positioning device (2) is an RTK-GNSS system. The advance speed of the harvester issued by the information processing system (3) is calculated according to the harvesting attribute of the rice to match the advance speed control model of the harvester, and is used to control the advance speed of the harvester; the advance speed control model is: V=λ1×D+λ2×H+C2; wherein V is the advance speed, D is the density of the rice in the ROI area, λ1 is the density proportionality coefficient, H is the plant height, λ2 is the plant height proportionality coefficient, and C2 is a constant; The harvesting attribute of the rice refers to the density and plant height of the rice. The density of the rice is calculated by a rice density calculation model; the rice density calculation model is: D=α×p+β×n+C1; wherein D is the density of the rice, p is the pixel area of the rice ear mask, n is the number of skeleton line nodes, α is the coefficient of p, β is the coefficient of n, and C1 is a constant. The method comprises the following steps: Step S1, image acquisition: the image acquisition device (1) is used to acquire the image of the rice in the area to be harvested on the left side of the harvester, obtain a color RGB image and a depth image, and transmit them to the information processing system; Step S2, position acquisition: the positioning device is used to measure the latitude and longitude coordinates and the heading of the current position of the harvester; Step S3, information processing: processing the rice image obtained in step S1 through the information processing system to obtain the plant height and density information of the rice, calculating the forward speed of the harvester through the control model; processing the harvester position and heading information to obtain the position information of the left side of the harvested area through the latitude and longitude coordinate conversion, and storing the rice harvesting attribute information and the position information one by one; when the harvester reaches a new position, the rice harvesting attribute information of the current harvesting position is called, the forward speed is calculated according to the forward speed control model, and the forward speed is transmitted to the forward speed control module (4); Step S4, speed control: the forward speed control module (4) controls the forward speed of the harvester according to the forward speed sent by step S3.

2. The control method of the rice harvesting property detection and work regulation system according to claim 1, characterized by, The image acquisition device (1) is a binocular camera that vertically downwardly photographs; the information processing system (3) is an embedded development board or an industrial computer.

3. The control method of the rice harvesting property detection and work regulation system according to claim 1, characterized by, The latitude and longitude coordinate conversion formula is: ； Wherein, (x1, y1) is the current position of the harvester latitude and longitude coordinates, (x2, y2) is the current camera harvester rice position latitude and longitude coordinates, R is the radius of the earth, L is the distance between the two positions, is the heading angle of the harvester, is the angle between the two position line and the heading of the harvester.

4. The control method of the rice harvesting property detection and work regulation system according to claim 1, characterized by, The rice density acquisition method comprises the following steps: According to the RGB image of the rice in the harvesting machine side of the harvested area collected by the image acquisition device (1), the ROI region is extracted with the harvesting machine cutting width as the scale; After the image in the ROI region is cut off, histogram equalization processing is performed, and then the processed image is input into a semantic segmentation network model to obtain a mask containing only the rice panicle; The pixel area of the rice panicle mask is calculated; The skeleton line extraction algorithm is used to extract the panicle skeleton line, and the number of skeleton line nodes is calculated; The rice density is calculated according to the rice density calculation model with the rice panicle pixel area and the number of skeleton line nodes as inputs and the rice density as output.

5. The control method of the rice harvesting property detection and work regulation system according to claim 1, characterized by, The rice plant height acquisition method comprises the following steps: By extracting the depth information of the rice surface to the camera in the ROI region, the average depth value h1 is obtained after filtering the abnormal values of the depth information through the abnormal value filtering method, and the difference between the camera installation height h2 is obtained, and the rice plant height H = h2-h1.

6. A harvester characterized by The control method of the rice harvesting attribute detection and operation control system according to any one of claims 1-5.

Images