LiDAR occlusion detection method and apparatus, storage medium, and LiDAR
The present application discloses a LiDAR occlusion detection method and apparatus, a storage medium, and a LiDAR. The method includes: obtaining detected echo data, obtaining distance information of each point in the echo data, comparing the distance information with a preset distance range, and in response to the distance information being within the preset distance range, determining that the LiDAR is occluded. In the present application, it can be detected in real time whether the LiDAR is occluded, without affecting transmittance of the LiDAR or increasing manufacturing costs of the LiDAR.
1 . A method for LiDAR occlusion detection, applied to a LiDAR, comprising:
when the LiDAR is occluded by a reference object, obtaining a reference distance between the reference object and the LiDAR, and obtaining a preset distance range based on the reference distance, wherein the reference distance refers to distance information of each point in echo data obtained by the LiDAR during normal working after the reference object is provided;
dividing a detection region into a general detection region and a target detection region, and setting alarm values based on a number of points within the preset distance range in the general detection region and the target detection region, wherein the target detection region is a detection region pertaining to a central region in the detection region of the LiDAR, the general detection region is a detection region other than the target detection region, and the alarm value set for the target detection region is higher than the alarm value for the general detection region;
obtaining detected echo data;
obtaining distance information of each point in the echo data;
comparing the distance information of each point in the echo data with the preset distance range;
when the distance information of any point is within the preset distance range, multiple pieces of echo data are further obtained for a region or a position in which the point is located;
if distance information of the multiple pieces of echo data of the point is within the preset distance range, determining that the LiDAR is occluded;
if only a preset proportion of the distance information of the multiple pieces of echo data is within the preset distance range, determining that the LiDAR is not occluded; and
obtaining a number of points whose distance information is within the preset distance range, and raising a fault alarm when the number of points within the preset distance range reaches the preset alarm value.
2 . The method of claim 1 , wherein before obtaining detected echo data, the method further comprises:
calculating a reference ratio of a reference number of points within the preset distance range in a detection region to a number of all points in the detection region, and based on the reference ratio, classifying different occlusion levels.
3 . The method of claim 1 , wherein after determining that the LiDAR is occluded, the method further comprises:
calculating a ratio of the number of points within the preset distance range in a detection region to a number of all points in the detection region, and based on the ratio and different preset occlusion levels, determining a current occlusion level; and
outputting the current occlusion level to a host computer, to enable the host computer to process corresponding to the current occlusion level.
4 . The method of claim 3 , wherein calculating a ratio of the number of points within the preset distance range in the detection region to the number of all points in the detection region comprises:
calculating a ratio of the number of points within the preset distance range in a target detection region to the number of all points in the target detection region; or
calculating a ratio of the number of points within the preset distance range in a general detection region to the number of all points in the general detection region; or
calculating a ratio of the number of points within the preset distance range in all detection regions to the number of all points in all the detection regions.