A Decomposition Model for Stereo Matching

TL;DR

This paper introduces a multi-scale decomposition model for stereo matching that significantly reduces computational costs by combining low-resolution dense matching with high-resolution sparse matching, achieving comparable accuracy with much faster processing.

Contribution

The proposed model innovatively decomposes stereo matching into multi-scale sparse and dense matching, reducing computational costs while maintaining accuracy.

Findings

Achieves 10-100x speed increase over state-of-the-art methods.

Maintains comparable disparity estimation accuracy.

Effectively fuses multi-scale disparity maps with occlusion awareness.

Abstract

In this paper, we present a decomposition model for stereo matching to solve the problem of excessive growth in computational cost (time and memory cost) as the resolution increases. In order to reduce the huge cost of stereo matching at the original resolution, our model only runs dense matching at a very low resolution and uses sparse matching at different higher resolutions to recover the disparity of lost details scale-by-scale. After the decomposition of stereo matching, our model iteratively fuses the sparse and dense disparity maps from adjacent scales with an occlusion-aware mask. A refinement network is also applied to improving the fusion result. Compared with high-performance methods like PSMNet and GANet, our method achieves $10-100\times$ speed increase while obtaining comparable disparity estimation results.