Towards urban scenes understanding through polarization cues

TL;DR

This paper introduces a polarization-based approach to improve urban scene understanding for autonomous robots, enhancing segmentation and depth estimation by leveraging light polarization cues in dynamic environments.

Contribution

It presents a novel two-axis pipeline utilizing polarization indices, demonstrating improved robustness and accuracy over RGB-centric methods in urban scene analysis.

Findings

Polarization cues improve segmentation accuracy.

Depth estimation quality is enhanced with polarization data.

The proposed method outperforms state-of-the-art RGB-based approaches.

Abstract

Autonomous robotics is critically affected by the robustness of its scene understanding algorithms. We propose a two-axis pipeline based on polarization indices to analyze dynamic urban scenes. As robots evolve in unknown environments, they are prone to encountering specular obstacles. Usually, specular phenomena are rarely taken into account by algorithms which causes misinterpretations and erroneous estimates. By exploiting all the light properties, systems can greatly increase their robustness to events. In addition to the conventional photometric characteristics, we propose to include polarization sensing. We demonstrate in this paper that the contribution of polarization measurement increases both the performances of segmentation and the quality of depth estimation. Our polarimetry-based approaches are compared here with other state-of-the-art RGB-centric methods showing interest of using polarization imaging.