期刊文献+

一种改进的实时半全局立体匹配算法及硬件实现 被引量:8

An Improved Real-Time Semi-Global Stereo Matching Algorithm and Its Hardware Implementation
下载PDF
导出
摘要 在基于现场可编程门阵列的实时立体匹配系统中,Census变换算法针对特定区域的误匹配率较高。为提高匹配精度,提出一种具有高并行性流水线结构的实时半全局立体匹配算法并进行硬件实现。将改进的Tanimoto距离和带权重4方向的梯度绝对值差进行组合,作为新的初始匹配代价。在代价聚合阶段采用4路径并行结构的SGM算法,在视差选择阶段采用赢家通吃策略,在视差校正阶段采用阈值检测算法代替传统左右一致性检验算法。实验结果表明,该算法能够有效提高弱纹理和边缘区域的区分度,减少对中心点的依赖,降低资源占用,其在Middleburry平台上的平均误匹配率仅为7.52%,在Xilinx Zynq-7000平台上的匹配速率达到98frame/s。 When applied to the real-time stereo matching systems based on Field Programmable Gate Array(FPGA),the Census Transform(CT)algorithm has a high false matching rate in specific areas.In order to improve the matching accuracy,a real-time Semi-Global stereo Matching(SGM)algorithm with highly parallel pipeline structure is proposed.The algorithm takes the combination of the improved Tanimoto distance and the weighted 4-direction Absolute Differences of Gradient(ADG)as the initial matching cost.In the cost aggregation stage,a 4-path SMG algorithm is used.In the disparity computation stage,the winner-takes-all strategy is chosen.Then in the parallax correction stage,the threshold detection algorithm is used to replace the traditional left-right check algorithm.Experimental results show that the proposed algorithm can effectively improve the discrimination between weak texture and edge regions,and reduce the dependence on center point and resource consumption.The average mismatch rate of the algorithm on Middlebury platform is 7.52%,and its matching rate on Xilinx zynq-7000 platform is 98 frame/s.
作者 赵晨园 李文新 张庆熙 ZHAO Chenyuan;LI Wenxin;ZHANG Qingxi(Lanzhou Institute of Physics,CAST,Lanzhou 730000,China)
出处 《计算机工程》 CAS CSCD 北大核心 2021年第9期162-170,共9页 Computer Engineering
基金 国家自然科学基金(61125101) 中国载人航天工程重大专项(RWZY640601)。
关键词 现场可编程门阵列 立体匹配 Census变换 Tanimoto距离 实时半全局匹配 Field Programmable Gate Array(FPGA) stereo matching Census Transform(CT) Tanimoto distance real-time Semi-Global stereo Matching(SGM)
  • 相关文献

参考文献7

二级参考文献41

  • 1陈大伟,刘海龙,李金屏.复杂静态背景下多移动目标实时检测系统的FPGA实现[J].吉林大学学报(工学版),2013,43(S1):287-290. 被引量:4
  • 2Stentoumis C, Grammatikopoulos L, Kalisperakis I, et al. A local adaptive approach for dense stereo matching in archi- tectural scene reconstruction[J]. International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sci- ences, 2013, 40(5W1): 219-226. 被引量:1
  • 3Gauglitz S, Hoellerer T, Turk M. Evaluation of interest point detectors and feature descriptors for visual tracking[J]. Interna- tional Journal of Computer Vision, 2011, 94(3): 335-360. 被引量:1
  • 4Humenberger M, Zinner C, Weber M, et al. A fast stereo match- ing algorithm suitable for embedded real-time system[J]. Com- puter Vision and Image Understanding, 2010, 114(11): 1180- 1202. 被引量:1
  • 5Scharstein D. Matching images by comparing their gradient fields[C]//IAPR International Conference on Pattern Recogni- tion. Piscataway, USA: IEEE, 1994: 572-575. 被引量:1
  • 6Fusiello A, Roberto V, Trucco E. Efficient stereo with multiple windowing[C]//IEEE Computer Society Conference on Com- puter Vision and Pattern Recognition. Piscataway, USA: IEEE, 1997: 858-863. 被引量:1
  • 7Yoon K J, Kweon I S. Adaptive support-weight approach for correspondence search[J]. IEEE Transactions on Pattern Analy- sis and Machine Intelligence, 2006, 28(4): 650-656. 被引量:1
  • 8Zhao Z S, Feng X, Teng S H, et al. Multiscale point correspon- dence using feature distribution and frequency domain align- ment[J]. Mathematical Problems in Engineering, 2012: No. 382369. 被引量:1
  • 9Zhao Z S, Feng X, Wei F, et al. Learning representative fea- tures for robot topological localization[J]. International Journal of Advanced Robotic System, 2013, 10:No.215. 被引量:1
  • 10Zhao Z S, Zhang L, Zhao M, et al. Gabor face recognition by multi-channel classifier fusion of supervised kernel manifold learning[J]. Neurocomputing, 2012, 97: 398-404. 被引量:1

共引文献101

同被引文献64

引证文献8

二级引证文献27

相关作者

内容加载中请稍等...

相关机构

内容加载中请稍等...

相关主题

内容加载中请稍等...

浏览历史

内容加载中请稍等...
;
使用帮助 返回顶部