Image-to-Point Cloud Registration Made Easy with Rectified Flow-based LiDAR Upsampling

May 1, 2026·
Reon Tabata
Reon Tabata
,
Kenji Koide
,
Shuji Oishi
,
Masashi Yokozuka
,
Taku Okawara
,
Aoki Takanose
,
Jun Miura
· 1 min read
Type
Publication
Submitted to the IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS 2026)
publications

This manuscript proposes converting sparse LiDAR point clouds into dense point-cloud images with a generative model, enabling image-domain feature extraction and matching methods to be applied to image-to-point-cloud registration. The manuscript was under review when the source document was prepared in May 2026.

LiDAR Computer Vision Multimodal Learning
Reon Tabata
Authors
Reon Tabata
M2 Student at Toyohashi University of Technology
Reon Tabata studies environmental perception and autonomous navigation for mobile robots. His work combines LiDAR, cameras, computer vision, and machine learning.