计算机科学 ›› 2021, Vol. 48 ›› Issue (1): 204-208.doi: 10.11896/jsjkx.191000205

• 计算机图形学与多媒体 • 上一篇    下一篇

基于PatchMatch的半全局高效双目立体匹配算法

桑苗苗1, 彭进先2, 达通航1, 张旭峰1   

  1. 1 中国人民解放军63618部队 新疆 库尔勒 841000
    2 中国人民解放军63611部队 新疆 库尔勒 841000
  • 收稿日期:2019-10-30 修回日期:2020-05-15 出版日期:2021-01-15 发布日期:2021-01-15
  • 通讯作者: 桑苗苗(1045699980@qq.com)

Efficient Semi-global Binocular Stereo Matching Algorithm Based on PatchMatch

SANG Miao-miao1, PENG Jin-xian2, DA Tong-hang1, ZHANG Xu-feng1   

  1. 1 Unit 63618 of PLA,Korla,Xinjiang 841000,China
    2 Unit 63611 of PLA,Korla,Xinjiang 841000,China
  • Received:2019-10-30 Revised:2020-05-15 Online:2021-01-15 Published:2021-01-15
  • About author:SANG Miao-miao,born in 1992,master,assistant engineer.Her main research interests include computer vision and so on.

摘要: 近年来双目立体匹配技术发展迅速,高精度、高分辨率、大视差的应用需求无疑对该技术的计算效率提出了更高的要求。由于传统立体匹配算法固有的计算复杂度正比于视差范围,已经难以满足高分辨率、大视差的应用场景。因此,从计算复杂度、匹配精度、匹配原理等多方面综合考虑,提出了一种基于PatchMatch的半全局双目立体匹配算法,在路径代价计算过程中使用空间传播机制,将可能的视差由整个视差范围降低为t个候选视差(t远远小于视差范围),显著减少了候选视差的数量,大幅提高了半全局算法的计算效率。对KITTI2015数据集的评估结果表明,该算法以5.81%的错误匹配率和20.2 s的匹配时间实现了准确性和实时性的明显提高。因此,作为传统立体匹配改进算法,该设计可以为大视差双目立体匹配系统提供高效的解决方案。

关键词: PatchMatch算法, 高精度大视差, 计算效率, 双目立体匹配

Abstract: In recent years,the binocular stereo matching has developed rapidly.The application of high accuracy,high resolution and large disparityput forward higher requirement for the computational efficiency.Since the computational complexity inherent in the traditional stereo matching algorithm is proportional to the disparity range,it has been difficult to meet the high resolution and large disparity applications.Considering the pros and cons of several types of stereo matching algorithms from the aspects of computational complexity,an efficient semi-global stereo matching algorithm based on PatchMatch through the effective combination of the two algorithms is proposed.It significantly reduces the computational complexity of the original SGM algorithm,since it reduces the possible disparity with only agroup of best t candidate disparities(t is much smaller than the disparity range) instead of the whole disparity range by means of the PatchMatch spatial propagation scheme.The evaluation results on KITTI2015 dataset demonstrate that the proposed algorithm achieves a significant improvement in accuracy and real-time performance with an 5.81% error matching rate and a matching time of 20.2 seconds.Therefore,as an improved algorithm for traditional stereo matching,this design can provide an efficient solution for large disparity binocular stereo matching system.

Key words: Binocular stereo matching, Computational efficiency, High accuracy and large disparity, PatchMatch algorithm

中图分类号: 

  • TP311.5
[1] WANG W T,HAN Z G,LIU P F.Multi-sensor Image Registration Algorithm Based on SIFT Points and Canny Edge Features Matching[J].Computer Science,2011,38(7):287-289.
[2] XU Z G,CHEN C.Robust and Fast Feature Points Matching[J].Computer Science,2013,40(2):294-296.
[3] MOZEROV M G,WEIJER J.Accurate Stereo Matching byTwo-Step Energy Minimization[J].IEEE Transactions on Ima-ge Processing,2015,24:1153-1163.
[4] ZHAN Y,GU Y,ZHANG K C,et al.Accurate image-guided stereo matching with efficient matching cost and disparity refinement[J].IEEE Transactions on Circuits and Systems for Video Technology,2015,99:11.
[5] HIRSCHMULLER H.Accurate and efficient stereo processing by semi-global matching and mutual information[C]//IEEE Conference on Computer Vision and Pattern Recognition.2005:807-814.
[6] HIRSCHMULLER H,SCHARSTEIN D.Evaluation of costfunctions forstereo matching[C]//IEEE Conference on Compu-ter Vision and Pattern Recognition.2007:1-8.
[7] KENDALL A,MARTIROSYAN H,DASGUPTA S,et al.End-to-End Learning of Geometry and Context for Deep Stereo Regression[C]//IEEE International Conference on Computer Vision.2017:66-75.
[8] ZBONTAR J,LECUN Y,et al.Stereo matching by training a convolutional neural network to compare image patches[J].Journal of Machine Learning Research,2016,17(2):1-32.
[9] SEKI A,POLLEFEYS M.SGM-Nets:Semi-global matchingwith neural networks[C]//IEEE Conference on Computer Vision and Pattern Recognition Workshops.2017:21-26.
[10] BARNES C,SHECHTMAN E,FINKELSTEIN A,et al.PatchMatch:A randomized correspondence algorithm for structural image editing[J].ACM Transactions on Graphics,2009,28(3):24-28.
[11] TTOFIS C,KYRKOU C,THEOCHARIDES T.A low-cost real-time embedded stereo vision system for accurate disparity estimation based on guided image filtering[J].IEEE Transactions on Computers,2016,65(9):2678-2693.
[12] CHEN S Q,ZHANG X C,SUN H B,et al.sWMF:Separable weighted median filter for efficient large-disparity stereo matching[C]//IEEE International Symposium on Circuits and Systems.2017:1-4.
[13] BLEYER M,RHEMANN C,ROTHER C.Patchmatch stereo-stereo matching with slanted support windows[C]//Machine Vision Conference.2011:1-11.
[14] BESSE F,ROTHER C,FITZGIBBON A W,et al.PMBP:Patchmatch belief propagation for correspondencefield estimation[J].International Journal of Computer Vision,2014,110(1):2-13.
[15] MENZE M,GEIGER A.Object scene flow for autonomous vehicles[C]//IEEE Conference on Computer Vision and Pattern Recognition.2015:3061-3070.
[16] SCHARSTEIN D,HIRSCHMÜLLER H,KITAJIMA Y,et al.High-resolution stereo datasets with subpixel accurate ground truth[C]//German Conference on Pattern Recognition.2014:31-42.
[17] AN W,FANGRONG W,BAICANG G,et al.Disparity map optimization based on edge detection[J].Computer Applications and Software,2019,36(7):236-241.
[1] 赵涛, 张凌浩, 赵其刚, 王红军.
基于聚类簇中心的共识跨链交换模型
Accross Block Chain Consensus Transation Model Based on Cluster Center
计算机科学, 2019, 46(11A): 557-561.
[2] 曹芳,朱永康.
基于SFS方法的三维重构及精度分析
3D Reconstruction Based on SFS Method and Accuracy Analysis
计算机科学, 2017, 44(Z6): 244-247. https://doi.org/10.11896/j.issn.1002-137X.2017.6A.056
Viewed
Full text


Abstract

Cited

  Shared   
  Discussed   
No Suggested Reading articles found!