Abstract:Superpixel provides local pixel coherence and respects object boundary, which is beneficial to stereo matching. Recently, superpixel cues are introduced into deep stereo networks. These methods develop a superpixel-based sampling scheme to downsample input color images and upsample output disparity maps. However, in this way, the image details are inevitably lost in the downsampling and the upsampling process introduces errors in the final disparity as well. Besides, this mechanism further limits the possibility of utilizing larger and multi-scale superpixels, which are important to alleviate the matching ambiguity. To address these problems, a superpixel-guided stereo matching method (LSG-Stereo) is proposed, which explicitly exploits the feature and disparity consistency within multi-scale superpixels to improve disparity estimation. To effectively incorporate superpixel cues into a stereo matching network, two novel modules are designed, including Superpixel Attention Spatial Pyramid Pooling (SA-SPP) and Superpixel-Guided Refinement (SGR). The SA-SPP module takes advantage of the content-aware superpixel pooling to construct an adaptive spatial pooling pyramid for better feature extraction. The SGR module explicitly utilizes the disparity consistency over multi-scale superpixels to further refine the disparity estimation in details and ill-posed regions. The proposed method is evaluated on the Scene Flow dataset, KITTI 2012, and KITTI 2015 stereo benchmarks with comprehensive experiments. Experimental results demonstrate that our method can significantly improve the accuracy of stereo matching, especially in details, occlusions, and texture-less regions.

Graph-based Stereo Matching by Incorporating Monocular Cues

Efficient Large Scale Stereo Matching Based on Cross-Scale.

Efficient Binocular Stereo Matching Based on Sad and Improved Census Transformation

Space-temporal stereo matching for dynamic scene

Dynamic Limited-Search Matching Algorithm Based on Geometric Properties of Stereo Imaging

Stereo Matching by Incorporating Depth Cues

Visually Imbalanced Stereo Matching

Superpixel Guided Network for Three-Dimensional Stereo Matching

Object segmentation using stereo images

Accurate Real-Time Stereo Correspondence Using Intra- and Inter-Scanline Optimization

Stereo Matching Using Graph Cuts: A 3D-Hough Transformation Approach.

Stereo matching from monocular images using feature consistency

Deep eyes: Joint depth inference using monocular and binocular cues

Superpixel-based Graph Cuts for Accurate Stereo Matching

A new combination of monocular and stereo cues for dense disparity estimation

A novel stereo matching pipeline with robustness and unfixed disparity search range

Disparity Estimation Based on the Combination of Depth Camera and Stereo Vision

Improving Stereo Matching by Incorporating Geometry Prior into Convnet

Extracting Disparity Map from Bifocal Monocular Stereo Vision in a Novel Way

Stereo Matching by Self-supervision of Multiscopic Vision.

MonStereo: When Monocular and Stereo Meet at the Tail of 3D Human Localization