JOURNAL ARTICLE

High Performance Implementation of Stereo Vision Matching Based on FPGA

Jingting DingXin DuWenhui ZhouJilin Liu

Year: 2011 Journal:   JOURNAL OF ELECTRONICS INFORMATION TECHNOLOGY Vol: 33 (3)Pages: 597-603   Publisher: Chinese Academy of Sciences, Institute of Electronics

Abstract

立体视觉系统在3维场景信息感知中起着重要的作用。其中立体匹配算法的运算复杂度较高,实时处理需要硬件实现匹配运算。但在现有的不多实现中,性能要求和硬件资源的矛盾突出。随着分辨率的增加,对处理速度和视差搜索范围都有更高的要求。对此,该文提出了一种立体匹配硬件实现结构,通过并行化算法子模块和合理安排流水结构来提高性能。匹配算法引入了自适应相关窗口的匹配策略,提升了深度不连续区域的视差质量。该方法结合左右一致性校验准则,可有效去除大部分错误匹配结果。整个匹配流程在单片现场可编程门阵列(FPGA)上实现,并在有限硬件资源条件下将视差搜索范围扩大到128像素。系统时钟60 MHz时,对于512512分辨率的立体图像,系统可以实现60帧/秒以上的处理速度。

Keywords:
Field-programmable gate array Computer science Stereopsis Matching (statistics) Computer vision Artificial intelligence Embedded system Mathematics

Metrics

0
Cited By
0.00
FWCI (Field Weighted Citation Impact)
2
Refs
0.24
Citation Normalized Percentile
Is in top 1%
Is in top 10%

Topics

Advanced Vision and Imaging
Physical Sciences →  Computer Science →  Computer Vision and Pattern Recognition
Robotics and Sensor-Based Localization
Physical Sciences →  Engineering →  Aerospace Engineering
Image and Video Stabilization
Physical Sciences →  Computer Science →  Computer Vision and Pattern Recognition
© 2026 ScienceGate Book Chapters — All rights reserved.