A Stereo Algorithm for Thin Obstacles and Reflective Objects

TL;DR

This paper introduces a novel stereo algorithm using trinocular and polarization data to better detect thin obstacles and reflective objects in outdoor environments, outperforming standard stereo methods.

Contribution

The paper presents a hierarchical disparity algorithm with polarization and semantic object triangulation, specifically designed for outdoor scenes with thin and reflective obstacles.

Findings

Reduced bad pixels to 9.27% from 18.4% in reflective scenes.

Detected 53% of wire pixels versus 5% by standard stereo.

Utilized trinocular and polarization data for improved obstacle detection.

Abstract

Stereo cameras are a popular choice for obstacle avoidance for outdoor lighweight, low-cost robotics applications. However, they are unable to sense thin and reflective objects well. Currently, many algorithms are tuned to perform well on indoor scenes like the Middlebury dataset. When navigating outdoors, reflective objects, like windows and glass, and thin obstacles, like wires, are not well handled by most stereo disparity algorithms. Reflections, repeating patterns and objects parallel to the cameras' baseline causes mismatches between image pairs which leads to bad disparity estimates. Thin obstacles are difficult for many sliding window based disparity methods to detect because they do not take up large portions of the pixels in the sliding window. We use a trinocular camera setup and micropolarizer camera capable of detecting reflective objects to overcome these issues. We present a hierarchical disparity algorithm that reduces noise, separately identify wires using semantic object triangulation in three images, and use information about the polarization of light to estimate the disparity of reflective objects. We evaluate our approach on outdoor data that we collected. Our method contained an average of 9.27% of bad pixels compared to a typical stereo algorithm's 18.4% of bad pixels in scenes containing reflective objects. Our trinocular and semantic wire disparity methods detected 53% of wire pixels, whereas a typical two camera stereo algorithm detected 5%.