High-resolution Stereo-vision Depth Processor with Min-pooling Method
High-resolution stereo-vision depth processor is widely adopted due to its cost-effectiveness and high information density. However, higher resolution demands significantly increase on hardware performance requirements. To address this challenge, the proposed Semi-global Matching (SGM) method with min-pooling effectively reduces memory usage by 50%. This optimization enables the achievement of state-of-the-art performance in both FPGA and ASIC implementations while minimizing memory resource consumption.
Authors
- Xianglong Wang (ORCID: https://orcid.org/0000-0003-1086-5158)
- Pingcheng Dong (ORCID: https://orcid.org/0000-0003-3975-286X)
- Fengwei An (ORCID: https://orcid.org/0000-0002-7554-7938)
- Gang Shi
- Lei Chen (ORCID: https://orcid.org/0000-0002-4334-6904)
- Wenyue Zhang (ORCID: https://orcid.org/0009-0002-6833-1626)
- Zhuoyu Chen (ORCID: https://orcid.org/0000-0003-0193-3534)
Institutions
- Hong Kong University of Science and Technology (HK)
- Southern University of Science and Technology (CN)
Publication Details
- Published
- 2024-10-25
- DOI
- https://doi.org/10.1109/icta64028.2024.10860544
- Primary Topic
- Advanced Vision and Imaging
- Type
- article
- Field-Weighted Citation Impact
- 0.00