Semantic 3D Mapping from Deep Image Segmentation

The perception and identification of visual stimuli from the environment is a fundamental capacity autonomous mobile robots. Current deep learning techniques make it possible to identify segment objects interest in an image. This paper presents novel algorithm object’s space segmentation image taken by 3D camera. proposed approach solves boundary pixel problem that appears when direct mapping segmented pixels their correspondence point cloud used. We validate our comparing baseline approaches using real images camera, showing method outperforms results terms accuracy reliability. As application algorithm, we present semantic for robot’s indoor environments.