LiDAR Moving Object Detection With Point Segmentation

The patent segments the point cloud data processing by separating moving objects from stationary background. It classifies measurement points into moving points and stationary points based on speed thresholds, then processes only the moving points for object detection. This segmentation reduces the volume of data requiring complex processing while maintaining detection accuracy for moving objects.

The patent extracts and removes stationary background points from the point cloud data, keeping only the moving points that represent actual objects of interest. By taking out the stationary components that consume processing resources but provide less useful information for dynamic scene understanding, the system reduces processing load while preserving detection capability for moving objects.

The patent performs preliminary classification of point cloud data into moving and stationary points before the main object detection process. By calculating speeds of all measurement points and classifying them in advance, the system prepares the data in a way that enables faster subsequent processing, reducing overall processing time while maintaining detection completeness for moving objects.

The patent applies partial processing by focusing computational resources only on moving points rather than processing all point cloud data equally. This partial action approach processes sufficient data (moving points) to achieve detection completeness while avoiding unnecessary processing of stationary background points, thereby reducing processing time.