Tail wave filtering method applied to unmanned ship

Abstract

The invention discloses a tail wave filtering method for unmanned ship application, and belongs to the field of unmanned ship water surface environment perception and understanding. The method comprises the following steps: s1, acquiring point cloud data of a water surface environment around the unmanned ship, wherein the point cloud data is obtained by scanning through a multi-line rotary laser radar; s2, carrying out clutter and noise removal on the point cloud data; s3, carrying out tail wave point identification on the filtered point cloud data, comprising: s31, calculating the actual height difference of reflection points of two adjacent laser beams in the vertical direction in the point cloud data, s32, calculating the ratio of the actual height difference to the theoretical height difference of each point cloud data point, and taking the ratio as a normalized height difference, and s33, identifying the point of which the normalized height difference is lower than a first threshold value as a tail wave point; s4, removing the tail wave point from the point cloud data. According to the method, the ratio of the actual height difference to the theoretical height difference of each point cloud data point serves as the normalized height difference feature, and the tail wave point cloud can be accurately recognized. Therefore, it is ensured that all tail wave points are found out, and then the tail wave point cloud is simply and efficiently removed.

Description

technical field [0001] The invention belongs to the field of perception and understanding of the water surface environment of unmanned boats, and more specifically relates to a wake wave filtering method for unmanned boats. Background technique [0002] Lidar is one of the most important environmental perception sensors for autonomous navigation of unmanned boats. The lidar periodically scans the surrounding water surface environment to obtain the azimuth and geometric shape information of the obstacle target. The unmanned boat uses the environment perception algorithm to identify and locate the obstacles in the surrounding environment, thereby completing autonomous obstacle avoidance navigation. However, when the ship is sailing on the water surface, the wake generated by the ship will reflect the laser and generate an echo point cloud in the lidar field of view, which is often misjudged by the environment perception algorithm as an obstacle on the water surface, resulting ...

AI Technical Summary

Problems solved by technology

One method is based on the assumption that the wake wave point cloud is at a relatively low height in the point cloud of the entire water surface scene, and uses the height threshold screening method to directly remove the point cloud with a low height in the entire scene; That part of the point cloud will also be removed, that is, "over-segmentation", resulting in "damage" of the target, thereby affecting the subsequent detection and recognition of the target, and in high sea conditions, the height range covered by the wake wave point cloud usually has a certain distance from the surface target. coincidence, the above assumptions are no longer satisfied, and the algorithm fails

One method is to perform voxel filtering or raster filtering on the entire scene point cloud, and calculate the voxel grid or raster The height difference of the point cloud in the grid removes the point cloud in the voxel grid or grid whose height difference is lower than a certain threshold; similarly, the simple threshold strategy is difficult to adapt to various sea conditions, and this method usually leaves some wake waves The point cloud is "under-segmented", and it will also be "over-segmented", removing part of the point cloud with a small local height difference in the target, which affects subsequent target detection and recognition, so this method is also ineffective

None of the existing methods can find features with strong classification ability for target points and wake points, and high scene adaptability, resulting in only partial wake filter effects being achieved in specific scenarios, resulting in "under-segmentation" or "over-segmentation". "segmentation" problem, failed to solve the problem of wake wave filtering

Images