Sparse cost volume for efficient stereo matching

Chuanhua Lu, Hideaki Uchiyama, Diego Thomas, Atsushi Shimada, Rin ichiro Taniguchi

Research output: Contribution to journalArticlepeer-review

29 Citations (Scopus)


Stereo matching has been solved as a supervised learning task with convolutional neural network (CNN). However, CNN based approaches basically require huge memory use. In addition, it is still challenging to find correct correspondences between images at ill-posed dim and sensor noise regions. To solve these problems, we propose Sparse Cost Volume Net (SCV-Net) achieving high accuracy, low memory cost and fast computation. The idea of the cost volume for stereo matching was initially proposed in GC-Net. In our work, by making the cost volume compact and proposing an efficient similarity evaluation for the volume, we achieved faster stereo matching while improving the accuracy. Moreover, we propose to use weight normalization instead of commonly-used batch normalization for stereo matching tasks. This improves the robustness to not only sensor noises in images but also batch size in the training process. We evaluated our proposed network on the Scene Flow and KITTI 2015 datasets, its performance overall surpasses the GC-Net. Comparing with the GC-Net, our SCV-Net achieved to: (1) reduce 73.08% GPU memory cost; (2) reduce 61.11% processing time; (3) improve the 3PE from 2.87% to 2.61% on the KITTI 2015 dataset.

Original languageEnglish
Article number1844
JournalRemote Sensing
Issue number11
Publication statusPublished - Nov 1 2018

All Science Journal Classification (ASJC) codes

  • General Earth and Planetary Sciences


Dive into the research topics of 'Sparse cost volume for efficient stereo matching'. Together they form a unique fingerprint.

Cite this