Search

CN-121661044-B - 2D laser point cloud detection method for right angle point of greenhouse intersection and storage medium

CN121661044BCN 121661044 BCN121661044 BCN 121661044BCN-121661044-B

Abstract

The invention belongs to the technical field of point cloud identification, and particularly relates to a 2D laser point cloud detection method of right angle points of a greenhouse intersection and a storage medium. The method comprises the steps of obtaining a single-frame laser point cloud of a robot 2D at a certain moment, establishing a cross sliding window, carrying out sliding detection on all points in the single-frame laser point cloud by the cross sliding window, detecting right angle points, constructing a point cloud set of the right angle points according to the detected right angle points, and calculating to obtain the optimal right angle points in the point cloud set. The method can effectively extract the right angle point from noise data, has better timeliness than an image Harris angle point detection algorithm, and consumes less than half of the time of an image method.

Inventors

  • WANG XIN
  • LIANG WEI
  • HAN JINFEN

Assignees

  • 山西工程科技职业大学

Dates

Publication Date
20260505
Application Date
20260204

Claims (8)

  1. 1. The 2D laser point cloud detection method for the right angle point of the greenhouse intersection is characterized by comprising the following steps of: Acquiring a single-frame laser point cloud of a robot 2D at a certain moment; Establishing a cross sliding window; the cross sliding window is divided into 5 areas from 0 to 4; wherein the 0 area is a corner area and is square and arranged in the middle; The 1-4 areas are channel edge areas and are rectangular, and are sequentially arranged on the right side, the upper side, the left side and the lower side of the 0 area; The cross sliding window is used for carrying out sliding detection on all points in the single-frame laser point cloud and detecting right angle points; The slip detection includes: starting sliding detection from the first point to the end of the last point; when the right angle point is positioned at the left lower corner of the right angle wall surface or at the right upper corner of the right angle wall surface, if Then the right angle point is considered to be detected; when the right angle point is positioned at the upper left corner of the right angle wall surface, if Then the right angle point is considered to be detected; Wherein, the The area code of the point outside the 0 area, And Representing the operations of extracting the maximum and minimum values respectively, The number of data in the set is represented, Representing a data set belonging to region 1, subscript j being the set belonging to ; Constructing a point cloud set of the right angle points according to the detected right angle points; And in the point cloud set, calculating to obtain an optimal right angle point.
  2. 2. The method for detecting 2D laser point cloud of right angle point of greenhouse intersection as claimed in claim 1, wherein the obtained single frame laser point cloud needs to rotate anticlockwise , Is the direction angle of the robot.
  3. 3. The method for detecting 2D laser point cloud of right angle point of greenhouse intersection according to claim 1, wherein the horizontal side length of the 2,0,4 area is the same as the vertical side length of the 1,0,3 area, and is 2l 2 .
  4. 4. The method for detecting 2D laser point cloud of right angle point of greenhouse crossing according to claim 3, , Is the standard deviation of the point cloud data.
  5. 5. The method for detecting 2D laser point cloud of right angle point of greenhouse intersection according to claim 1, wherein half of long side of cross sliding window is l 1 ,l 1 smaller than channel width w.
  6. 6. The method for detecting 2D laser point clouds at right angles to a greenhouse intersection according to claim 1, wherein if the number of point clouds in a certain area of 1 to 4 areas is smaller than a set threshold, the statistical value of the number of point clouds in the area is set to 0.
  7. 7. The method for 2D laser point cloud detection of right angle points at a greenhouse intersection according to claim 1, wherein calculating an optimal right angle point comprises: when the right angle point is positioned at the lower left corner of the right angle wall surface, a minimum objective function of the mixing distance is established: Wherein the method comprises the steps of Is an objective function of this case; The number of the areas 0 and 2 except the point i in the anticlockwise direction of the point i; the number of the areas 0 and 1 except the point i in the clockwise direction of the point i; when the right angle point is positioned at the right lower corner of the right angle wall surface, a minimum value objective function of the mixing distance is established: Wherein the method comprises the steps of Is an objective function of this case; the number of the areas 0 and 3 except the point i in the anticlockwise direction of the point i; the number of the areas 0 and 2 except the point i in the clockwise direction of the point i; When the right angle point is positioned at the right upper corner of the right angle wall surface, a minimum value objective function of the mixing distance is established: Wherein the method comprises the steps of Is an objective function of this case; The number of the areas 0 and 4 except the point i in the anticlockwise direction of the point i; The number of the areas 0 and 3 except the point i in the clockwise direction of the point i; when the right angle point is positioned at the upper left corner of the right angle wall surface, a minimum objective function of the mixing distance is established: Wherein the method comprises the steps of Is an objective function of this case; the number of the areas 0 and 1 except the point i in the anticlockwise direction of the point i; The number of the areas 0 and 4 except the point i in the clockwise direction of the point i; And solving the minimum objective function to obtain the optimal right angle point.
  8. 8. A computer readable storage medium, on which a computer program is stored, characterized in that the computer program, when being executed by a processor, implements a 2D laser point cloud detection method of right angle points of a greenhouse intersection according to any one of claims 1-7.

Description

2D laser point cloud detection method for right angle point of greenhouse intersection and storage medium Technical Field The invention belongs to the technical field of point cloud identification, and particularly relates to a 2D laser point cloud detection method of right angle points of a greenhouse intersection and a storage medium. Background Many right angle points are arranged in crossing channels of places such as greenhouses, and the identification of right angle points in single-frame 2D laser point clouds has important significance for real-time position estimation of robots at crossings. For the extraction of the corner feature, the line segment feature is generally extracted first, and then the intersection point of the two line segments is calculated to obtain the corner feature. In addition, there is an image method, which needs to convert the point cloud into a picture, then identify the corner points in the picture by using the Harris corner point detection algorithm, and finally needs to convert the pixel coordinates of the corner points into actual coordinates. The prior method has complicated steps, can not directly extract the right angle points of the point cloud, and has low extraction efficiency. The existing method is influenced by straight line fitting precision or image resolution, and is influenced by the noise of a laser radar sensor, so that the right angle point recognition accuracy is low. Disclosure of Invention The invention aims to solve the problems and provides a 2D laser point cloud detection method for right angle points of a greenhouse intersection and a storage medium. The invention adopts the following technical scheme that the 2D laser point cloud detection method for the right angle point of the greenhouse intersection comprises the following steps: Acquiring a single-frame laser point cloud of a robot 2D at a certain moment; A cross sliding window is established; The cross sliding window is used for carrying out sliding detection on all points in the single-frame laser point cloud and detecting right angle points; constructing a point cloud set of the right angle points according to the detected right angle points; And in the point cloud set, calculating to obtain an optimal right angle point. In some embodiments, the acquired single frame laser point cloud needs to be rotated counterclockwise,Is the direction angle of the robot. In some embodiments, the cross-slide window is divided into 5 regions of 0 to 4; wherein the 0 area is a corner area and is square and arranged in the middle; The 1-4 areas are channel edge areas and are rectangular, and are sequentially arranged on the right side, the upper side, the left side and the lower side of the 0 area. In some embodiments, the horizontal side length of the 2,0,4 region and the vertical side length of the 1,0,3 region are the same, both being 2l 2. In some embodiments of the present invention, in some embodiments,,Is the standard deviation of the point cloud data. In some embodiments, half of the long side of the cross-shaped window is l 1,l1 less than the channel width w. In some embodiments, the slide detection includes: starting sliding detection from the first point to the end of the last point; when the right angle point is positioned at the left lower corner of the right angle wall surface or at the right upper corner of the right angle wall surface, if Then the right angle point is considered to be detected; when the right angle point is positioned at the upper left corner of the right angle wall surface, if Then the right angle point is considered to be detected; Wherein, the The area code of the point outside the 0 area,AndRepresenting the operations of extracting the maximum and minimum values respectively,The number of data in the set is represented,Representing a data set belonging to region 1, subscript j being the set belonging to。 In some embodiments, if the number of point clouds in a certain area of the 1 to 4 areas is smaller than the set threshold, the statistical value of the number of point clouds in the area is set to 0. In some embodiments, computing the optimal corner point comprises: when the right angle point is positioned at the lower left corner of the right angle wall surface, a minimum objective function of the mixing distance is established: Wherein the method comprises the steps of Is an objective function of this case; The number of the areas 0 and 2 except the point i in the anticlockwise direction of the point i; the number of the areas 0 and 1 except the point i in the clockwise direction of the point i; the right angle point is positioned at the right lower corner of the right angle wall surface, and a minimum objective function of the mixing distance is established: Wherein the method comprises the steps of Is an objective function of this case; the number of the areas 0 and 3 except the point i in the anticlockwise direction of the point i; the number of the areas 0 and 2