What question did this study set out to answer?

This work aims to create accurate 3D urban maps with semantic labeling using street view images and LiDAR data.

Urban 3D segmentation and modelling from street view images and LiDAR point clouds

View Full Paper

Ask AI

Bookmark

Authors

Randomized trial investigates 3D urban mapping using images and point clouds, suggesting improved accuracy for autonomous systems.

This work aims to create accurate 3D urban maps with semantic labeling using street view images and LiDAR data.
Utilized GPS registered LiDAR point clouds and street view images as inputs.
Implemented a hybrid approach combining rule-based parsing and learning-based labeling for segmentation.
Adopted boosted decision trees for complex structures like cars and trees, exploiting both LiDAR and photometric features.
Achieved high segmentation accuracy for large structures, covering 75% of point cloud data.
Demonstrated improved 3D visualization and 2D image segmentation by back projecting 3D labels onto street view images.
Compared favorably with competing methods on three popular databases: NAVTEQ True, Paris-Rue-Madame, and TLS Velodyne.

Babahajiani et al. (2017) studied this question.

View Full Paper

Ask AI

Bookmark