Small Obstacle Avoidance Based on RGB-D Semantic Segmentation

TL;DR

This paper introduces a new RGB-D based obstacle avoidance system for road robots that uses a two-stage semantic segmentation network and morphological processing to accurately detect small obstacles and plan safe paths.

Contribution

The paper proposes a novel two-stage semantic segmentation network combined with morphological processing for precise obstacle detection in road robots.

Findings

High accuracy in indoor and outdoor obstacle detection

Effective avoidance of small obstacles missed by previous methods

Improved temporal consistency with optical flow supervision

Abstract

This paper presents a novel obstacle avoidance system for road robots equipped with RGB-D sensor that captures scenes of its way forward. The purpose of the system is to have road robots move around autonomously and constantly without any collision even with small obstacles, which are often missed by existing solutions. For each input RGB-D image, the system uses a new two-stage semantic segmentation network followed by the morphological processing to generate the accurate semantic map containing road and obstacles. Based on the map, the local path planning is applied to avoid possible collision. Additionally, optical flow supervision and motion blurring augmented training scheme is applied to improve temporal consistency between adjacent frames and overcome the disturbance caused by camera shake. Various experiments are conducted to show that the proposed architecture obtains high performance both in indoor and outdoor scenarios.