Mobile Object Route Control Using Height-Based Light Noise Filtering

The patent segments the point group data by spatial location (dividing the detection area into multiple regions) and by height direction (analyzing distribution along the vertical axis). This segmentation allows the system to apply noise determination processing to specific regions where light source noise is likely to occur, rather than processing all points uniformly, thereby improving noise recognition accuracy while controlling computational complexity.

The patent introduces height direction analysis as an additional dimension for noise detection. By examining the distribution of points along the vertical axis and identifying multiple maximum points in this dimension, the system can distinguish light source noise (which appears as detached points in the height direction) from actual obstacles. This dimensional approach enhances noise recognition without requiring complex algorithmic changes.

The patent applies noise determination processing selectively to specific regions rather than uniformly across the entire detection area. By identifying regions where light source noise is likely to occur (such as areas with detached points in the height direction) and applying processing only to those regions, the system maintains high route determination accuracy while reducing overall processing time and computational load.

The patent performs noise determination processing on a partial basis - specifically on point groups in regions where noise is suspected based on height direction distribution characteristics. This partial processing approach removes sufficient noise to ensure route determination accuracy while avoiding the computational overhead of processing all points, thus balancing reliability and productivity.