縮退環境下における欠損のあるLiDARの反射強度画像を使用したフォトメトリック誤差最小化に基づく姿勢補正の検証

抄録

This report presents a validation of a pose correction method based on minimizing photometric error in LiDAR reflectance intensity images, even when these images contain missing data due to low-density point clouds in degenerate environments. Previous approaches have used reflectance intensity images generated from LiDAR point clouds to correct the pose; however, these methods typically require high-density point clouds, and significant missing data can occur when the density is low. In this work, we address that limitation by employing a photometric error minimization approach, enabling robust pose correction even in the presence of incomplete reflectance intensity images. The effectiveness of the proposed method was verified using data acquired in a laboratory environment.