A global path planning method for unmanned vehicle operations based on DSM images
By filtering and image biclassifying the digital surface model of the tobacco field, an accurate tobacco field road network is generated, which solves the efficiency of the operation path planning of unmanned vehicles in the tobacco field, and achieves full coverage and distance reduction of the operation area.
Patent Information
- Application Number
- CN202410512565.1
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2024-04-26
- Publication Date
- 2024-12-06
- Estimated Expiration
- 2044-04-26
AI Technical Summary
In the tobacco field, it is difficult for the existing technology to effectively plan the operation path of unmanned vehicles, so that unmanned vehicles can minimize the operating distance while achieving full coverage of the operation area.
By obtaining the DSM image of the digital surface model of Yantian, performing maximum filtering and minimum filtering processing, obtaining the difference filtering image, and using the Otsu algorithm to perform image binary classification, refine operations and vectorization processing, generate Yantian Road Network, thereby planning The global operation path of the unmanned vehicle.
The accuracy of the Yantian Road Network has been improved, allowing unmanned vehicles to operate along the tobacco ridge, reduce operating distance, and achieve full coverage of the operation area.
Smart Images

Figure CN118362128B_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of image processing, and in particular to a global path planning method for unmanned vehicle operations based on DSM images. Background Art
[0002] As an intelligent agricultural machinery, unmanned vehicles can operate autonomously in tobacco fields according to preset paths, which can improve the efficiency and quality of tobacco production and reduce production costs. One of the core technologies of unmanned vehicles is path planning. The purpose of path planning is to achieve full coverage of the operation area while minimizing the operation distance. The operation method of tobacco fields is ridge operation, that is, the unmanned vehicle needs to operate along the ridges of tobacco, rather than randomly moving. How to reasonably plan the operation path of unmanned vehicles in tobacco fields so that the unmanned vehicle can minimize the operation distance while achieving full coverage of the operation area is an urgent problem to be solved. Summary of the invention
[0003] The purpose of the present invention is to provide a global path planning method for unmanned vehicle operations based on DSM images, which is used to reasonably plan the operation path of the unmanned vehicle in the tobacco field so that the unmanned vehicle can minimize the operation distance while achieving full coverage of the operation area.
[0004] According to the present invention, a global path planning method for unmanned vehicle operation based on DSM images comprises the following steps:
[0005] S100, obtaining a tobacco field digital surface model DSM image.
[0006] S200, obtaining a maximum value filtering image and a minimum value filtering image of a tobacco field, wherein the maximum value filtering image and the minimum value filtering image of the tobacco field are obtained by performing maximum value filtering processing and minimum value filtering processing on a DSM image, respectively.
[0007] S300, obtaining a tobacco field difference filter image, wherein the tobacco field difference filter image is obtained by subtracting a tobacco field maximum value filter image and a tobacco field minimum value filter image.
[0008] S400, using the Otsu algorithm to perform binary classification on the tobacco field difference filter image to obtain a binary image of the tobacco field; in the binary image of the tobacco field, pixels in the ridge and ditch area are assigned a value of 1, and pixels in the ridge and platform area are assigned a null value.
[0009] S500, performing a thinning operation on the binary image of the tobacco field, and converting the raster after the thinning operation into a vector line.
[0010] S600, performs topological analysis on the vector lines, connects intersecting lines, deletes hanging lines, and smoothes the vector lines to obtain a tobacco field road network.
[0011] S700: Plan a global operation path for the unmanned vehicle according to the tobacco field road network.
[0012] The present invention has at least the following beneficial effects:
[0013] The present invention obtains a tobacco field digital surface model DSM image, performs a subtraction between a tobacco field maximum value filtered image obtained by performing maximum value filtering processing on the DSM image and a tobacco field minimum value filtered image obtained by performing minimum value filtering processing on the DSM image, obtains a tobacco field difference filtered image, and then uses an Otsu algorithm to perform image binary classification on the tobacco field filtered image, to obtain a tobacco field binary image; the present invention obtains the tobacco field binary image based on the tobacco field difference filtered image, can solve the problem of inaccurate tobacco field binary image obtained by directly using the Otsu algorithm to perform image binary classification on the DSM image due to the fact that some areas in the tobacco field are higher than other areas, and improves the accuracy of the tobacco field road network finally determined. Moreover, after obtaining the binary image of the tobacco field, the present invention obtains a tobacco field road network by performing a refinement operation on the binary image, converting the raster to vector lines, connecting intersecting lines, deleting hanging lines and smoothing the vector lines. The lines constituting the tobacco field road network are the roads in the tobacco field that the unmanned vehicle can travel on. Planning the global operation path of the unmanned vehicle according to the tobacco field road network can enable the unmanned vehicle to operate along the tobacco ridges instead of walking randomly, which is beneficial for the unmanned vehicle to minimize the operating distance while achieving full coverage of the operating area. BRIEF DESCRIPTION OF THE DRAWINGS
[0014] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.
[0015] Figure 1 A flow chart of a global path planning method for unmanned vehicle operation based on DSM images provided in an embodiment of the present invention;
[0016] Figure 2 A schematic diagram of a tobacco field road network provided by an embodiment of the present invention;
[0017] Figure 3 A schematic diagram of an operation area provided by an embodiment of the present invention;
[0018] Figure 4 A schematic diagram of a path for an unmanned vehicle to travel from Pr'2 to zPr'2 and a path for an unmanned vehicle to travel from Pr'1 to zPr'1 provided in an embodiment of the present invention;
[0019] Figure 5The unmanned vehicle provided in the embodiment of the present invention travels from xPr'2 to P 1,1 A schematic diagram of the path;
[0020] Figure 6 The unmanned vehicle provided in the embodiment of the present invention travels from xPr'2 to P 1,2 A schematic diagram of the path;
[0021] Figure 7 The unmanned vehicle provided in the embodiment of the present invention travels from Pr'1 to P' 1,1 A schematic diagram of the path;
[0022] Figure 8 The unmanned vehicle provided in the embodiment of the present invention travels from Pr'1 to P' 1,2 A schematic diagram of the path;
[0023] Fig. 9 A schematic diagram of a path of an unmanned vehicle within an operating area provided in an embodiment of the present invention. DETAILED DESCRIPTION
[0024] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative work are within the scope of protection of the present invention.
[0025] According to the present invention, a global path planning method for unmanned vehicle operation based on DSM images is provided, such as Figure 1 As shown, including:
[0026] S100, obtaining a tobacco field digital surface model DSM image.
[0027] Those skilled in the art are aware that any method for obtaining a DSM image in the prior art falls within the protection scope of the present invention.
[0028] S200, obtaining a maximum value filtering image and a minimum value filtering image of a tobacco field, wherein the maximum value filtering image and the minimum value filtering image of the tobacco field are obtained by performing maximum value filtering processing and minimum value filtering processing on a DSM image, respectively.
[0029] Those skilled in the art know that the processes of maximum value filtering and minimum value filtering are prior arts and will not be described in detail here.
[0030] Preferably, using a filter window that is a square with a side length of twice the plant spacing can ensure that the filter window located on the center line of the ridge contains at least one tobacco plant, making the filtering result of the ridge center line position more continuous, thereby better distinguishing the ridge and ditch positions.
[0031] S300, obtaining a tobacco field difference filter image, wherein the tobacco field difference filter image is obtained by subtracting a tobacco field maximum value filter image and a tobacco field minimum value filter image.
[0032] S400, using the Otsu algorithm to perform binary classification on the tobacco field difference filter image to obtain a binary image of the tobacco field; in the binary image of the tobacco field, pixels in the ridge and ditch area are assigned a value of 1, and pixels in the ridge and platform area are assigned a null value.
[0033] Those skilled in the art are aware that any method in the prior art for obtaining a binary image using the Otsu algorithm falls within the protection scope of the present invention.
[0034] S500, performing a thinning operation on the binary image of the tobacco field, and converting the raster after the thinning operation into a vector line.
[0035] Those skilled in the art are aware that any method of performing thinning operations on a binary image in the prior art falls within the protection scope of the present invention.
[0036] S600, performs topological analysis on the vector lines, connects intersecting lines, deletes hanging lines, and smoothes the vector lines to obtain a tobacco field road network.
[0037] As a specific implementation method, the tobacco field road network is as follows Figure 2 shown.
[0038] Those skilled in the art know that any smoothing method in the prior art falls within the protection scope of the present invention.
[0039] S700: Plan a global operation path for the unmanned vehicle according to the tobacco field road network.
[0040] In this embodiment, S700 further includes:
[0041] S710, setting the unmanned vehicle operation starting position P on the tobacco field road network s , end position P e and operating areas.
[0042] As a specific implementation method, the unmanned vehicle operation starting position is P s , the end position of the unmanned vehicle operation is P e , the operation area is the area within the dotted box, such as Figure 3 shown.
[0043] S720, planning a path for the unmanned vehicle within the operation area.
[0044] In this embodiment, S720 includes:
[0045] S721, obtaining the intersection of the working area and the tobacco field road network, R = (r1, r2, ..., r i ,...,r n ), r i It is the i-th intersecting working section between the working area and the tobacco field road network, the value range of i is 1 to n, and n is the number of intersecting working sections between the working area and the tobacco field road network.
[0046] Preferably, before S721, S720 also includes determining whether the ratio of the area of the unmanned vehicle operation area to the area of the tobacco field is less than a preset threshold. If so, then proceed to S721; if not, then plan the global path of the unmanned vehicle operation as follows: s Starting from the working area, the distance P s Drive into the work area from the nearest intersection work section end point, work in the work area using an S-shaped path, and drive out of the work area to reach P after the work is completed. e . Preferably, the preset threshold is greater than or equal to 0.5. In this embodiment, the ratio of the unmanned vehicle operation area to the tobacco field area is less than the preset threshold as a condition for executing S721-S726. When the unmanned vehicle operation area accounts for a large proportion, the S-type path operation mode can be directly adopted, thereby saving computing power without significantly increasing the length of the operation path. When the unmanned vehicle operation area accounts for a small proportion, S721-S726 can be adopted to significantly reduce the length of the operation path, thereby saving operation time and improving operation efficiency.
[0047] S722, obtaining the first side endpoint Pr'1 and the second side endpoint Pr'2 of the initial road segment r', where r' is a certain r in R i .
[0048] In this embodiment, r' is an intersecting working section in R selected randomly or an intersecting working section in R selected according to a preset condition.
[0049] Optionally, the first side endpoint is an endpoint located on the right side of the intersecting working section in the tobacco field road network, and the second side endpoint is an endpoint located on the left side of the intersecting working section in the tobacco field road network, such as Figure 4 shown.
[0050] S723, obtaining the endpoint zPr'1 closest to Pr'1 and the endpoint zPr'2 closest to Pr'2 among the endpoints of the initial remaining intersecting working sections; the initial remaining intersecting working sections are the other intersecting working sections in R except r'.
[0051] In this embodiment, if a certain endpoint P0 is a first-side endpoint, the distance from P0 to Pr'1 is the length of the shortest path connecting P0 and Pr'1 on the first side in the tobacco field road network; if a certain endpoint P0 is a second-side endpoint, the distance from P0 to Pr'1 is the sum of the distance between another endpoint (i.e., Pr'2) on the same intersecting working section as Pr'1 and P0 and the length of the intersecting working section (i.e., r') where Pr'1 is located, and the distance between Pr'2 and P0 is the sum of the shortest path connecting Pr'2 and P0 on the second side in the tobacco field road network. Length; if an endpoint P0 is an endpoint on the first side, the distance from P0 to Pr'2 is the sum of the distance between another endpoint (i.e. Pr'1) on the same intersecting working section as Pr'2 and P0 and the length of the intersecting working section (i.e. r') where Pr'2 is located, and the distance between Pr'1 and P0 is the length of the shortest path connecting Pr'1 and P0 on the first side in the tobacco field road network; if an endpoint P0 is an endpoint on the second side, the distance from P0 to Pr'2 is the length of the shortest path connecting P0 and Pr'2 on the second side in the tobacco field road network.
[0052] S724, if the path length of the unmanned vehicle from Pr'1 to zPr'1 is greater than the path length of the unmanned vehicle from Pr'2 to zPr'2, then enter S725; if the path length of the unmanned vehicle from Pr'1 to zPr'1 is less than or equal to the path length of the unmanned vehicle from Pr'2 to zPr'2, then enter S726.
[0053] In this embodiment, if zPr'2 is the second side endpoint, the path of the unmanned vehicle from Pr'2 to zPr'2 is the shortest path connecting zPr'2 and Pr'2 on the second side in the tobacco field road network, such as Figure 4 As shown by the thick black line on the left; if zPr'2 is the endpoint on the first side, the path of the unmanned vehicle from Pr'2 to zPr'2 is the path of the unmanned vehicle first driving from Pr'2 to Pr'1, and then driving from Pr'1 to zPr'2. The path from Pr'1 to zPr'2 is the shortest path connecting zPr'2 and Pr'1 on the first side in the tobacco field road network. If zPr'1 is the endpoint on the first side, the path of the unmanned vehicle from Pr'1 to zPr'1 is the shortest path connecting zPr'1 and Pr'1 on the first side in the tobacco field road network, as shown in Figure 4 As shown by the thick black line on the right side; if zPr'1 is the endpoint on the second side, the path of the unmanned vehicle from Pr'1 to zPr'1 is the path where the unmanned vehicle first drives from Pr'1 to Pr'2, and then drives from Pr'2 to zPr'1. The path from Pr'2 to zPr'1 is the shortest path connecting zPr'1 and Pr'2 on the second side in the tobacco field road network.
[0054] S725, if zPr'2 is the endpoint on the second side, the path planned for the unmanned vehicle to travel from Pr'1 to xPr'2 is Pr'1-Pr'2-zPr'2-xPr'2 or the path planned for the unmanned vehicle to travel from xPr'2 to Pr'1 is xPr'2-zPr'2-Pr'2-Pr'1, and xPr'2 is the other endpoint of the intersecting working section where zPr'2 is located; if zPr'2 is the endpoint on the first side, the path planned for the unmanned vehicle to travel from Pr'2 to xPr'2 is Pr'2-Pr'1-zPr'2-xPr'2 or the path planned for the unmanned vehicle to travel from xPr'2 to Pr'2 is xPr'2-zPr'2-Pr'1-Pr'2, and xPr'2 is the other endpoint of the intersecting working section where zPr'2 is located.
[0055] In this embodiment, in S725, if zPr'2 is the second side endpoint, the first side endpoint P of the intersecting operation section closest to r' in the first remaining intersecting operation section is also obtained. 1,1 and the second side endpoint P 1,2 , and enter S7251; the first remaining intersecting operation sections are other intersecting operation sections in R except the intersecting operation sections where r' and zPr'2 are located.
[0056] S7251, if the unmanned vehicle travels from Pr'1 to P 1,1 The path length is greater than the path length from Pr'1 to P 1,2 The path length is xPr'2, so the unmanned vehicle is planned to travel from xPr'2 to P 1,1 The path is xPr'2-zPr'2-Pr'2-Pr'1-Pr'2-P 1,2 -P 1,1 Or planning unmanned vehicles from P 1,1 The path to xPr'2 is P 1,1 -P 1,2 -Pr'2-Pr'1-Pr'2-zPr'2-xPr'2; if the unmanned vehicle travels from Pr'1 to P 1,1 The path length is less than or equal to the distance from Pr'1 to P 1,2 The path length is xPr'2, so the unmanned vehicle is planned to travel from xPr'2 to P 1,2 The path is: xPr'2-zPr'2-Pr'2-Pr'1-P 1,1 -P 1,2 Or planning unmanned vehicles from P 1,2 The path to xPr'2 is: P 1,2 -P 1,1 -Pr'1-Pr'2-zPr'2-xPr'2.
[0057] In this embodiment, the unmanned vehicle travels from xPr'2 to P1,1 The path is xPr'2-zPr'2-Pr'2-Pr'1-Pr'2-P 1,2 -P 1,1 ,like Figure 5 The black thick line shows that the unmanned vehicle travels from xPr'2 to P 1,2 The path is xPr'2-zPr'2-Pr'2-Pr'1-P 1,1 -P 1,2 ,like Figure 6 Shown in bold black line.
[0058] In this embodiment, in S725, if zPr'2 is the first side endpoint, the first side endpoint P of the intersecting operation section closest to r' in the first remaining intersecting operation section is also obtained. 1,1 and the second side endpoint P 1,2 , and enter S7252; the first remaining intersecting operation sections are other intersecting operation sections in R except the intersecting operation sections where r' and zPr'2 are located.
[0059] S7252, if the unmanned vehicle travels from Pr'2 to P 1,1 The path length is greater than the path length from Pr'2 to P 1,2 The path length is xPr'2, so the unmanned vehicle is planned to travel from xPr'2 to P 1,1 The path is xPr'2-zPr'2-Pr'1-Pr'2-P 1,2 -P 1,1 Or planning unmanned vehicles from P 1,1 The path to xPr'2 is P 1,1 -P 1,2 -Pr'2-Pr'1-zPr'2-xPr'2; if the unmanned vehicle travels from Pr'2 to P 1,1 The path length is less than or equal to the distance from Pr'2 to P 1,2 The path length is xPr'2, so the unmanned vehicle is planned to travel from xPr'2 to P 1,2 The path is: xPr'2-zPr'2-Pr'1-Pr'2-Pr'1-P 1,1 -P 1,2 Or planning unmanned vehicles from P 1,2 The path to xPr'2 is: P 1,2 -P 1,1 -Pr'1-Pr'2-Pr'1-zPr'2-xPr'2.
[0060] In this embodiment, in S725, if zPr'2 is the second side endpoint, the first side endpoint P' of the intersecting operation section closest to the intersecting operation section where zPr'2 is located in the first remaining intersecting operation section is also obtained. 1,1and the second side endpoint P' 1,2 , and enter S7253.
[0061] S7253, if the unmanned vehicle travels from xPr'2 to P' 1,1 The path length is greater than the path length from the unmanned vehicle traveling from xPr'2 to P' 1,2 The path length is, then the unmanned vehicle is planned to travel from Pr'1 to P' 1,1 The path is Pr'1-Pr'2-zPr'2-xPr'2-zPr'2-P 1,2 -P' 1,1 Or planning unmanned vehicles from P' 1,1 The path to Pr'1 is P' 1,1 -P 1,2 -zPr'2-xPr'2-zPr'2-Pr'2-Pr'1; if the unmanned vehicle travels from xPr'2 to P' 1,1 The path length is less than or equal to the distance from xPr'2 to P' 1,2 The path length is, then the unmanned vehicle is planned to travel from Pr'1 to P' 1,2 The path is Pr'1-Pr'2-zPr'2-xPr'2-P' 1,1 -P' 1,2 Or planning unmanned vehicles from P' 1,2 The path to Pr'1 is P' 1,2 -P' 1,1 -xPr'2-zPr'2-Pr'2-Pr'1.
[0062] As a specific implementation, the unmanned vehicle travels from Pr'1 to P' 1,1 The path is Pr'1-Pr'2-zPr'2-xPr'2-zPr'2-P 1,2 -P' 1,1 ,like Figure 7 The black thick line shows that the unmanned vehicle travels from Pr'1 to P' 1,2 The path is Pr'1-Pr'2-zPr'2-xPr'2-P' 1,1 -P' 1,2 ,like Figure 8 Shown in bold black line.
[0063] In this embodiment, based on S7251, it is determined that the vehicle travels from Pr'1 to P 1,1 The path length is greater than the path length from Pr'1 to P 1,2 The path length of the unmanned vehicle is xPr'2 and P 1,1 The driving path between Pr'1 and Pr'2 is determined based on S7251. 1,1The path length is less than or equal to the distance from Pr'1 to P 1,2 The path length of the unmanned vehicle is xPr'2 and P 1,2 Based on S7253, it is determined that the driving path from xPr'2 to P' 1,1 The path length is greater than the path length from the unmanned vehicle traveling from xPr'2 to P' 1,2 The path length of the unmanned vehicle is between Pr'1 and P' 1,1 The driving path between the two, and the unmanned vehicle driving from xPr'2 to P' based on S7253 1,1 The path length is less than or equal to the distance from xPr'2 to P' 1,2 The path length of the unmanned vehicle is between Pr'1 and P' 1,2 The driving path between. This embodiment can realize the path planning of the unmanned vehicle including 2 intersecting working sections in the working area based on S725, and can further realize the path planning of the unmanned vehicle including 4 intersecting working sections in the working area based on S7251 and S7253. According to a similar method, the two endpoints of the determined path in the working area and the remaining intersecting working sections are continuously updated, and then the endpoints closest to the two endpoints of the latest determined path are respectively found in the remaining intersecting working sections. According to methods similar to S7251 and S7253, the path in the working area can be continuously updated until all the intersecting working paths in the working area are planned; if in the process of planning the path of the unmanned vehicle in the working area, the intersecting working section where one of the two endpoints of the latest determined path is located is r1 or r n , then the n The endpoint is a path endpoint of the operation area. There is no need to update this endpoint. Just update the other endpoint in a similar way to S7251 or S7253 until the intersection operation section where the other endpoint is located is also r1 or r n , then the other endpoint is the other path endpoint of the working area.
[0064] S726, if zPr'1 is the endpoint on the first side, the path planned for the unmanned vehicle to travel from Pr'2 to xPr'1 is Pr'2-Pr'1-zPr'1-xPr'1 or the path planned for the unmanned vehicle to travel from xPr'1 to Pr'2 is xPr'1-zPr'1-Pr'1-Pr'2, and xPr'1 is the other endpoint of the intersecting working section where zPr'1 is located; if zPr'1 is the endpoint on the second side, the path planned for the unmanned vehicle to travel from Pr'1 to xPr'1 is Pr'1-Pr'2-zPr'1-xPr'1 or the path planned for the unmanned vehicle to travel from xPr'1 to Pr'1 is xPr'1-zPr'1-Pr'2-Pr'1, and xPr'1 is the other endpoint of the intersecting working section where zPr'1 is located.
[0065] Based on S721-S726 of this embodiment, the operating distance of the unmanned vehicle in the operating area is minimized, and the operating path of the unmanned vehicle in the tobacco field is reasonably planned, so that the unmanned vehicle can minimize the operating distance while achieving full coverage of the operating area, thereby improving the operating efficiency.
[0066] S730, obtaining two path endpoints of the unmanned vehicle in the operation area.
[0067] S740, if P s To P a The path and P b To P e The sum of the paths is less than or equal to P s To P b The path and P a To P e The sum of the paths, then P s To P a The path and operation area P a To P b The path and P b To P e The path is determined as the global path of the unmanned vehicle; if P s To P a The path and P b To P e The sum of the paths is greater than P s To P b The path and P a To P e The sum of the paths, then P s To P b The path and operation area P b To P a The path and P a To P e The path is determined as the global path of the unmanned vehicle; ais the first path endpoint of the unmanned vehicle in the operating area, P b It is the second path endpoint of the unmanned vehicle in the operating area.
[0068] In this embodiment, P s To P a The path is the path connecting P s With P a The shortest path, P s To P b The path is the path connecting P s With P b The shortest path, P b To P e The path is the path connecting P b With P e The shortest path, P a To P e The path is the path connecting P a With P e The shortest path.
[0069] In this embodiment, based on S720, only the operation path of the unmanned vehicle in the operation area can be determined, and the driving direction of the unmanned vehicle cannot be determined; if P s To P a The path and operation area P a To P b The path and P b To P e The path is determined as the global path of the unmanned vehicle, then P a is the starting point in the operation area, then the driving direction of the unmanned vehicle in the operation area is determined; if P s To P b The path and operation area P b To P a The path and P a To P e The path is determined as the global path of the unmanned vehicle, then P b It is the starting point in the working area, so the driving direction of the unmanned vehicle in the working area is determined; in the two cases, the driving directions of the unmanned vehicle in the working area are opposite, but the driving paths are the same.
[0070] As a specific implementation method, the path of the unmanned vehicle in the operation area is as follows: Fig. 9 As shown by the thick black line, P a and P b like Fig. 9 shown.
[0071] This embodiment obtains a digital surface model DSM image of a tobacco field, and subtracts a maximum value filtered image of the tobacco field obtained by performing maximum value filtering on the DSM image from a minimum value filtered image of the tobacco field obtained by performing minimum value filtering on the DSM image to obtain a difference filtered image of the tobacco field, and then uses the Otsu algorithm to perform image binary classification on the filtered image of the tobacco field to obtain a binary image of the tobacco field. This embodiment obtains a binary image of the tobacco field based on the difference filtered image of the tobacco field, which can solve the problem of inaccurate binary images of the tobacco field obtained by directly using the Otsu algorithm to perform image binary classification on the DSM image due to the fact that some areas in the tobacco field are higher than other areas, thereby improving the accuracy of the tobacco field road network finally determined. Moreover, after obtaining the binary image of the tobacco field, this embodiment obtains the tobacco field road network by performing refinement operations on the binary image, converting raster to vector lines, connecting intersecting lines, deleting hanging lines and smoothing the vector lines. The lines constituting the tobacco field road network are the roads in the tobacco field that the unmanned vehicle can travel on. Planning the global operation path of the unmanned vehicle according to the tobacco field road network can enable the unmanned vehicle to operate along the tobacco ridges instead of walking randomly, which is conducive to minimizing the operating distance of the unmanned vehicle while achieving full coverage of the operating area.
[0072] Although some specific embodiments of the present invention have been described in detail by way of example, it will be appreciated by those skilled in the art that the above examples are for illustration only and are not intended to limit the scope of the present invention. It will also be appreciated by those skilled in the art that various modifications may be made to the embodiments without departing from the scope and spirit of the present invention. The scope of the present invention is defined by the appended claims.
Claims
1. A global path planning method for unmanned vehicle operation based on DSM image, characterized in that: The following steps are involved: S100, obtaining a digital surface model DSM image of a tobacco field; S200, obtaining a maximum value filtering image and a minimum value filtering image of a tobacco field, wherein the maximum value filtering image and the minimum value filtering image of a tobacco field are obtained by performing maximum value filtering and minimum value filtering on a DSM image, respectively; S300, obtaining a tobacco field difference filter image, wherein the tobacco field difference filter image is obtained by subtracting a tobacco field maximum value filter image and a tobacco field minimum value filter image; S400, using the Otsu algorithm to perform image binary classification on the tobacco field difference filter image to obtain a tobacco field binary image; in the tobacco field binary image, pixels in the ridge and ditch area are assigned a value of 1, and pixels in the ridge and platform area are assigned a value of null; S500, performing a thinning operation on the binary image of the tobacco field, and converting the raster after the thinning operation into a vector line; S600, topological analysis is performed on the vector lines, intersecting lines are connected, hanging lines are deleted, and the vector lines are smoothed to obtain a tobacco field road network; S700, planning a global operation path for the unmanned vehicle according to the tobacco field road network; The S700 also includes: S710, setting the unmanned vehicle operation starting position P on the tobacco field road network s , end position P e and operating areas; S720, planning a path for the unmanned vehicle in the operation area; S730, obtaining two path endpoints of the unmanned vehicle in the operation area; S740, if P s To P a The path and P b To P e The sum of the paths is less than or equal to P s To P b The path and P a To P e The sum of the paths, then P s To P a The path and operation area P a To P b The path and P b To P e The path is determined as the global path of the unmanned vehicle; if P s To P a The path and P b To P e The sum of the paths is greater than P s To P b The path and P a To P e The sum of the paths, then P s To P b The path and operation area P b To P a The path and P a To P e The path is determined as the global path of the unmanned vehicle; a is the first path endpoint of the unmanned vehicle in the operating area, P b is the second path endpoint of the unmanned vehicle in the operation area; The S720 includes: S721, obtaining the intersection of the working area and the tobacco field road network, R = (r1, r2, ..., r i ,...,r n ), r i is the i-th intersecting working section between the working area and the tobacco field road network, i is in the range of 1 to n, and n is the number of intersecting working sections between the working area and the tobacco field road network; S722, obtaining the first side endpoint Pr'1 and the second side endpoint Pr'2 of the initial road segment r', where r' is a certain r in R i ; S723, obtaining the endpoint zPr'1 closest to Pr'1 and the endpoint zPr'2 closest to Pr'2 among the endpoints of the initial remaining intersecting working sections; the initial remaining intersecting working sections are the other intersecting working sections in R except r'; S724, if the path length of the unmanned vehicle from Pr'1 to zPr'1 is greater than the path length of the unmanned vehicle from Pr'2 to zPr'2, then enter S725; if the path length of the unmanned vehicle from Pr'1 to zPr'1 is less than or equal to the path length of the unmanned vehicle from Pr'2 to zPr'2, then enter S726; S725, if zPr'2 is the endpoint of the second side, the planned path for the unmanned vehicle to travel from Pr'1 to xPr'2 is Pr'1-Pr'2-zPr'2-xPr'2 or the planned path for the unmanned vehicle to travel from xPr'2 to Pr'1 is xPr'2-zPr'2-Pr'2-Pr'1, and xPr'2 is the other endpoint of the intersecting working section where zPr'2 is located; if zPr'2 is the endpoint of the first side, the planned path for the unmanned vehicle to travel from Pr'2 to xPr'2 is Pr'2-Pr'1-zPr'2-xPr'2 or the planned path for the unmanned vehicle to travel from xPr'2 to Pr'2 is xPr'2-zPr'2-Pr'1-Pr'2; S726, if zPr'1 is the endpoint of the first side, the planned path for the unmanned vehicle to travel from Pr'2 to xPr'1 is Pr'2-Pr'1-zPr'1-xPr'1 or the planned path for the unmanned vehicle to travel from xPr'1 to Pr'2 is xPr'1-zPr'1-Pr'1-Pr'2, and xPr'1 is the other endpoint of the intersecting working section where zPr'1 is located; if zPr'1 is the endpoint of the second side, the planned path for the unmanned vehicle to travel from Pr'1 to xPr'1 is Pr'1-Pr'2-zPr'1-xPr'1 or the planned path for the unmanned vehicle to travel from xPr'1 to Pr'1 is xPr'1-zPr'1-Pr'2-Pr'1.
2. The global path planning method for unmanned vehicle operation based on DSM image according to claim 1 is characterized in that: In S725, if zPr'2 is the second side endpoint, the first side endpoint P of the intersecting working section closest to r' in the first remaining intersecting working section is also obtained. 1,1 and the second side endpoint P 1,2 , and enter S7251; the first remaining intersecting operation sections are other intersecting operation sections in R except the intersecting operation sections where r' and zPr'2 are located; S7251, if the unmanned vehicle travels from Pr'1 to P 1,1 The path length is greater than the path length from Pr'1 to P 1,2 The path length is xPr'2, so the unmanned vehicle is planned to travel from xPr'2 to P 1,1 The path is xPr'2-zPr'2-Pr'2-Pr'1-Pr'2-P 1,2 -P 1,1 Or planning unmanned vehicles from P 1,1 The path to xPr'2 is P 1,1 -P 1,2 -Pr'2-Pr'1-Pr'2-zPr'2-xPr'2; if the unmanned vehicle travels from Pr'1 to P 1,1 The path length is less than or equal to the distance from Pr'1 to P 1,2 The path length is xPr'2, so the unmanned vehicle is planned to travel from xPr'2 to P 1,2 The path is: xPr'2-zPr'2-Pr'2-Pr'1-P 1,1 -P 1,2 Or planning unmanned vehicles from P 1,2 The path to xPr'2 is: P 1,2 -P 1,1 -Pr'1-Pr'2-zPr'2-xPr'2.
3. The global path planning method for unmanned vehicle operation based on DSM image according to claim 1 is characterized in that: In S725, if zPr'2 is the first side endpoint, the first side endpoint P of the intersecting working section closest to r' in the first remaining intersecting working section is also obtained. 1,1 and the second side endpoint P 1,2 , and enter S7252; the first remaining intersecting operation sections are other intersecting operation sections in R except the intersecting operation sections where r' and zPr'2 are located; S7252, if the unmanned vehicle travels from Pr'2 to P 1,1 The path length is greater than the path length from Pr'2 to P 1,2 The path length is xPr'2, so the unmanned vehicle is planned to travel from xPr'2 to P 1,1 The path is xPr'2-zPr'2-Pr'1-Pr'2-P 1,2 -P 1,1 Or planning unmanned vehicles from P 1,1 The path to xPr'2 is P 1,1 -P 1,2 -Pr'2-Pr'1-zPr'2-xPr'2; if the unmanned vehicle travels from Pr'2 to P 1,1 The path length is less than or equal to the distance from Pr'2 to P 1,2 The path length is xPr'2, so the unmanned vehicle is planned to travel from xPr'2 to P 1,2 The path is: xPr'2-zPr'2-Pr'1-Pr'2-Pr'1-P 1,1 -P 1,2 Or planning unmanned vehicles from P 1,2 The path to xPr'2 is: P 1,2 -P 1,1 -Pr'1-Pr'2-Pr'1-zPr'2-xPr'2.
4. The global path planning method for unmanned vehicle operation based on DSM image according to claim 1 is characterized in that: In S725, if zPr'2 is the second side endpoint, the first side endpoint P' of the intersecting operation section closest to the intersecting operation section where zPr'2 is located in the first remaining intersecting operation section is also obtained. 1,1 and the second side endpoint P' 1,2 , and enter S7253; the first remaining intersecting operation sections are other intersecting operation sections in R except the intersecting operation sections where r' and zPr'2 are located; S7253, if the unmanned vehicle travels from xPr'2 to P' 1,1 The path length is greater than the path length from the unmanned vehicle traveling from xPr'2 to P' 1,2 The path length is, then the unmanned vehicle is planned to travel from Pr'1 to P' 1,1 The path is Pr'1-Pr'2-zPr'2-xPr'2-zPr'2-P 1,2 -P' 1,1 Or planning unmanned vehicles from P' 1,1 The path to Pr'1 is P' 1,1 -P 1,2 -zPr'2-xPr'2-zPr'2-Pr'2-Pr'1; if the unmanned vehicle travels from xPr'2 to P' 1,1 The path length is less than or equal to the distance from xPr'2 to P' 1,2 The path length is, then the unmanned vehicle is planned to travel from Pr'1 to P' 1,2 The path is Pr'1-Pr'2-zPr'2-xPr'2-P' 1,1 -P' 1,2 Or planning unmanned vehicles from P' 1,2 The path to Pr'1 is P' 1,2 -P' 1,1 -xPr'2-zPr'2-Pr'2-Pr'1.
5. The global path planning method for unmanned vehicle operation based on DSM image according to claim 1 is characterized in that: Before S721, S720 also includes determining whether the ratio of the area of the unmanned vehicle operation area to the area of the tobacco field is less than a preset threshold. If so, then enter S721.
6. The global path planning method for unmanned vehicle operation based on DSM image according to claim 5 is characterized in that: The preset threshold is greater than or equal to 0.5.
