Obstacle avoidance laser detection device point cloud power line detection method

Abstract

The invention relates to a point cloud power line detection method for an obstacle avoidance laser detection device, and belongs to the technical field of laser detection. The invention provides a point cloud power line detection method of an obstacle avoidance laser detection device, aiming at the detection and perception requirements of an external environment under a complex environment and a severe visual condition under a motion condition, and by utilizing the technical characteristics of a laser three-dimensional imaging detection mode, the obstacle avoidance auxiliary decision-making technology research is carried out, and the invention provides the point cloud power line detection method of the obstacle avoidance laser detection device. The method is a point cloud power line detection method based on a laser detection device, the power line is detected through active three-dimensional imaging detection, the obstacle avoidance requirement under the motion condition can be met, the power line detection capacity is effectively improved, the stable and reliable environment and situation awareness capacity can be provided for personnel, and safety is guaranteed.

Description

Technical Field

[0001] The present invention belongs to the technical field of laser detection, and in particular relates to a point cloud power line detection method for an obstacle avoidance laser detection device. Background Art

[0002] In order to meet the needs of detecting and perceiving the external environment in complex environments and harsh visual conditions under motion conditions, it is necessary to carry out research on obstacle avoidance decision-making assistance technology and design a point cloud power line detection method for an obstacle avoidance laser detection device. Summary of the invention

[0003] 1. Technical issues to be resolved

[0004] The technical problem to be solved by the present invention is: to meet the obstacle avoidance requirements, a power line detection method for an obstacle avoidance laser detection device is proposed.

[0005] (II) Technical solution

[0006] In order to solve the above technical problems, the present invention provides a point cloud power line detection method of an obstacle avoidance laser detection device, comprising the following steps:

[0007] Step 1: Extract valid laser point cloud data and remove infinitely far points in the original point cloud data;

[0008] Step 2: Crop the original point cloud through a straight-through filter to remove points without echoes, and perform elevation filtering based on the height to remove ground points;

[0009] Step 3, voxel filtering based on horizontal extension: calculate the cuboid that just wraps the point cloud, divide the cuboid into small flat cuboids of different heights, calculate the centroid of the small cuboid, and use the coordinates of the centroid to approximate several points in the cuboid, and set a threshold for the number of coordinate points on this basis. When the number of points in the cuboid is less than the threshold, filter out the point;

[0010] Step 4: Power line detection based on local features: traverse each point in the point cloud P i, construct the neighborhood Ni of point Pi and calculate the center of the neighborhood And the neighborhood covariance matrix, and calculate the eigenvalue of the point Ni in the point cloud, determine the main components of the point Ni according to the importance, and separate the power lines and other objects;

[0011] Step 5, generating a bounding box of a three-dimensional point cloud: first, calculate the covariance matrix of the point cloud, diagonalize the covariance matrix, calculate the eigenvalues, and orthogonally normalize the corresponding three eigenvectors. The three orthogonally normalized eigenvectors obtained are the three coordinate axis directions of the bounding box of the power line;

[0012] Step 6. After obtaining the three coordinate axis directions of the bounding box, all three-dimensional point clouds of the target are represented by these three coordinate axes, and the projection values ​​of the point clouds on the three coordinate axes are calculated. The length, width, height, and center coordinate point of the bounding box are obtained, and the point cloud power line detection is completed.

[0013] Preferably, in step 1, a laser detection device is used to collect point cloud, and the attitude angle information is obtained through an IMU attitude sensor, point cloud accumulation is performed, and then effective laser point cloud data is extracted.

[0014] Preferably, the laser detection device operates in two scanning modes: elliptical scanning and azimuth swing scanning.

[0015] The present invention also provides a system for implementing the method.

[0016] The present invention also provides an obstacle avoidance auxiliary decision method implemented based on the method.

[0017] (III) Beneficial effects

[0018] Aiming at the needs of detecting and perceiving the external environment in complex environments and harsh visual conditions under motion conditions, the present invention uses the technical characteristics of laser three-dimensional imaging detection to carry out obstacle avoidance auxiliary decision-making technology research, and proposes a point cloud power line detection method for obstacle avoidance laser detection device. This method is a point cloud power line detection method based on a laser detection device. Through active three-dimensional imaging detection, power lines are detected, which can meet the obstacle avoidance needs under motion conditions, effectively improve the power line detection capability, and provide personnel with stable and reliable environmental and situational awareness capabilities to ensure safety. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] Figure 1 It is a flow chart of point cloud power line detection of the present invention;

[0020] Figure 2 This is the experimental scene diagram;

[0021] Figure 3 is the original point cloud image;

[0022] Figure 4 is the extracted power line point;

[0023] Figure 5 is the bounding box of the 3D point cloud. DETAILED DESCRIPTION

[0024] In order to make the purpose, content and advantages of the present invention more clear, the specific implementation methods of the present invention are further described in detail below in conjunction with the drawings and examples.

[0025] The present invention proposes a three-dimensional laser point cloud obstacle detection method based on partition elevation filtering and principal component analysis to extract typical obstacles such as electric wires to reduce the risk of collision with obstacles under moving conditions.

[0026] The main technical means of the present invention are:

[0027] 1. Build a laser detection device point cloud imaging system: Use a laser detection device to collect point clouds, and use the IMU attitude sensor to obtain attitude angle information for point cloud accumulation.

[0028] 2. Establish a point cloud power line detection algorithm model: crop the point cloud through a straight-through filter, use the elevation filtering method to filter out ground points to improve the processing speed, use the horizontal extension-based voxel filtering technology to filter and downsample the point cloud, extract power lines based on local features of the processed point cloud, and use the principal component analysis method to generate bounding boxes to complete the point cloud power line detection.

[0029] 3. Experimental analysis of point cloud power line detection algorithm: Through experimental analysis, the performance of the point cloud power line detection method in different situations is verified, proving that the method can meet the obstacle avoidance requirements.

[0030] The laser detection device used in the present invention adopts an elliptical scanning + azimuth scanning two-dimensional scanning mode, with a field of view of 41.2°×30°, a maximum detection distance of 1500m, a point cloud quantity of 100kp / s, and a maximum detection distance of 800m for power lines.

[0031] Figure 1 This is a power line detection flow chart. The specific implementation process of this method is as follows:

[0032] Step 1: Extract valid laser point cloud data and remove infinitely far points in the original point cloud data;

[0033] Step 2: Crop the original point cloud through a straight-through filter to remove points without echoes, and perform elevation filtering based on the height to remove ground points;

[0034] Step 3, voxel filtering based on horizontal extension: calculate the cuboid that just wraps the point cloud, divide the cuboid into small flat cuboids of different heights, calculate the centroid of the small cuboid, and use the coordinates of the centroid to approximate several points in the cuboid, and set a threshold for the number of coordinate points on this basis. When the number of points in the cuboid is less than the threshold, filter out the point;

[0035] Step 4: Power line detection based on local features: traverse each point Pi in the point cloud, build the neighborhood Ni of point Pi, and calculate the center of the neighborhood And the neighborhood covariance matrix, and calculate the eigenvalue of the point Ni in the point cloud, determine the main components of the point Ni according to the importance, and separate the power lines and other objects;

[0036] Step 5, generating a bounding box of a three-dimensional point cloud: first, calculate the covariance matrix of the point cloud, diagonalize the covariance matrix, calculate the eigenvalues, and orthogonally normalize the corresponding three eigenvectors. The three orthogonally normalized eigenvectors obtained are the three coordinate axis directions of the bounding box of the power line;

[0037] Step 6. After obtaining the three coordinate axis directions of the bounding box, all three-dimensional point clouds of the target are represented by these three coordinate axes, and the projection values ​​of the point clouds on the three coordinate axes are calculated. The length, width, height, and center coordinate point of the bounding box are obtained, and the point cloud power line detection is completed.

[0038] The following specific cases are used to verify the effectiveness of the point cloud power line detection method. Figure 2 The scene image detected by the laser detection device, the point cloud data after point cloud collection is as follows Figure 3 As shown in Figure 2, the points after power line extraction are as follows: Figure 4 As shown, Figure 5 The point cloud image after the bounding box is generated is shown. Experiments show that this method can quickly and accurately detect power lines in laser point cloud data, has good noise resistance and robustness, and has certain application value under motion conditions.

[0039] The above is only a preferred embodiment of the present invention. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the technical principles of the present invention. These improvements and modifications should also be regarded as the scope of protection of the present invention.

Claims

1. A point cloud power line detection method for an obstacle avoidance laser detection device, characterized in that: The following steps are involved: Step 1: Extract valid laser point cloud data and remove infinitely far points in the original point cloud data; Step 2: Crop the original point cloud through a straight-through filter to remove points without echoes, and perform elevation filtering based on the height to remove ground points; Step 3, voxel filtering based on horizontal extension: calculate the cuboid that just wraps the point cloud, divide the cuboid into small flat cuboids of different heights, calculate the centroid of the small cuboid, and use the coordinates of the centroid to approximate several points in the cuboid, and set a threshold for the number of coordinate points on this basis. When the number of points in the cuboid is less than the threshold, filter out the point; Step 4: Power line detection based on local features: traverse each point Pi in the point cloud, build the neighborhood Ni of point Pi, and calculate the center of the neighborhood And the neighborhood covariance matrix, and calculate the eigenvalue of the point Ni in the point cloud, determine the main components of the point Ni according to the importance, and separate the power lines and other objects; Step 5, generating a bounding box of a three-dimensional point cloud: first, calculate the covariance matrix of the point cloud, diagonalize the covariance matrix, calculate the eigenvalues, and orthogonally normalize the corresponding three eigenvectors. The three orthogonally normalized eigenvectors obtained are the three coordinate axis directions of the bounding box of the power line; Step 6. After obtaining the three coordinate axis directions of the bounding box, all three-dimensional point clouds of the target are represented by these three coordinate axes, and the projection values ​​of the point clouds on the three coordinate axes are calculated. The length, width, height, and center coordinate point of the bounding box are obtained, and the point cloud power line detection is completed.

2. The method according to claim 1, characterized in that In step 1, a laser detection device is used to collect point cloud data, and the attitude angle information is obtained through the IMU attitude sensor to accumulate point cloud data, and then extract valid laser point cloud data.

3. The method according to claim 2, characterized in that The laser detection device works in two scanning modes: elliptical scanning and azimuth swing scanning.

4. The method according to claim 3, characterized in that The field of view of the laser detection device is 41.2°×30°, and the maximum detection distance is 1500m.

5. The method according to any one of claims 1 to 4, characterized in that The method is applied in the field of laser detection technology.

6. A system for implementing the method according to any one of claims 1 to 5.

7. The system according to claim 6, characterized in that The system is used in the field of laser detection technology.

8. An obstacle avoidance auxiliary decision method implemented based on the method as described in any one of claims 1 to 5.

Images