基于自适应权重的立体匹配优化算法

doi:10.19678/j.issn.1000-3428.0056991

摘要/Abstract

摘要： 为解决现有立体匹配算法对低纹理以及视差不连续区域匹配效果较差的问题，提出一种改进的立体匹配优化算法。在传统自适应权重算法匹配代价的基础上，融合高斯差分图像差分信息，即左右图像高斯差分图的差分，重新定义其初始匹配代价，增加算法在视差不连续区域的鲁棒性，并加入边缘约束和视差边缘约束迭代聚类以及基于高斯差分图的自适应窗口算法，保证改进算法在低纹理区域的匹配性能，消除坏点与视差空洞。将该算法与传统自适应权重匹配算法分别在Middlebury数据集上进行匹配实验，结果表明，该算法平均性能提升了15.05%，明显优于传统自适应权重匹配算法。

关键词: 立体匹配, 自适应权重, 低纹理, 视差不连续, 边缘约束, 自适应窗口

Abstract: In order to solve the problem that the existing stereo matching algorithm has poor performance for low texture and disparity-discontinuous regions, this paper proposes an improved stereo matching optimization algorithm. Based on the matching cost of the traditional adaptive weighting algorithm, the difference information of the Gaussian difference image (referring to the difference between the left and right image Gaussian difference map) is fused, and the initial matching cost is redefined to enhance the robustness of the algorithm for disparity-discontinuous regions. Then it adds edge constraint, parallax edge constraint iterative clustering and an adaptive window algorithm based on Gaussian difference map to ensure the matching performance of the algorithm in low-texture regions and eliminate dead pixels and parallax holes. The algorithm and the traditional adaptive weight matching algorithm are tested on the Middlebury dataset. The results show that the average performance of the proposed algorithm is improved by 15.05%, which is significantly better than that of the traditional adaptive weight matching algorithm.

Key words: stereo matching, adaptive weight, low texture, disparity discontinuity, edge constraint, adaptive window

中图分类号:

TP391

文斌, 朱晗. 基于自适应权重的立体匹配优化算法[J]. 计算机工程, 2021, 47(4): 268-276.

WEN Bin, ZHU Han. Stereo Matching Optimization Algorithm Based on Adaptive Weights[J]. Computer Engineering, 2021, 47(4): 268-276.

http://www.ecice06.com/CN/Y2021/V47/I4/268

图/表 12

20210425171647

20210425171653

20210425171656

20210425171659

20210425171701

20210425171704

20210425171707

20210425171710

20210425171713

20210425171715

20210425171718

20210425171721

参考文献

[1] GONZALEZ A,GARRIDO M A,LIODCA D F,et al.Automatic traffic signs and panels inspection system using computer vision[J].IEEE Transactions on Intelligent Transportation Systems,2011,12(2):485-499.
[2] CHEN Aihua,GAO Chenghui,HE Bingwei.Stereo-vision-enabled image stereo-matching method[J].Sponsored by China Construction Machinery,2014,12(3):194-198.(in Chinese)陈爱华,高诚辉,何炳蔚.立体视觉中的图像立体匹配方法[J].中国工程机械学报,2014,12(3):194-198.
[3] GEIGER A,ROSER M,URTASUM R.Efficient large-scale stereo matching[C]//Proceedings of ACCV'10.Queenstown.Wellington,New Zealand:[s.n.],2010:25-38.
[4] SCHARSTEIN D.A taxonomy and evaluation of dense two-frame stereo correspondence algorithms[C]//Proceedings of IEEE Workshop on Stereo and Multi-baseline Vision.Kauai,USA:[s.n.],2001:154-166.
[5] TRUCCO E,VERRI A.Introductory techniques for 3-D computer vision[M].[S.1.]:Prentice Hall,1998.
[6] KANG S B,SZELISKI R,CGAI J.Handling occlusions in dense multi-view stereo[C]//Proceedings of IEEE Computer Society Conference on Computer Vision and Pattern Recognition.Washington D.C.,USA:IEEE Press,2001:103-110.
[7] WU Q,LAW R.The forecasting model based on modified SVRM and PSO penalizing Gaussian noise[J].Expert Systems with Applications,2011,38(3):1887-1894.
[8] XU Yizhen,WANG Yongcheng,SHEN Zhou.A fast algorithm for matching multiple patterns[J].Journal of Shanghai Jiao Tong University,2002,36(4):516-520.(in Chinese)许一震,王永成,沈洲.一种快速的多模式字符串匹配算法[J].上海交通大学学报,2002,36(4):516-520.
[9] YOON K J,KWEON I S.Adaptive support-weight approach for correspondence search[J].IEEE Transactions on Pattern Analysis & Machine Intelligence,2006,28(4):650-566.
[10] CHEN D,ARDABILIAN M,CHEN L,et al.A fast trilateral filter-based adaptive support weight method for stereo matching[J].IEEE Transactions on Circuits and Systems for Video Technology,2015,25(5):730-743.
[11] LIM J,PARK J,PARK H.Principal depth estimation using cost-volume filtering for fast stereo matching[C]//Proceedings of 2018 International Conference on Elec-tronics,Information,and Communication.Honolulu,USA:[s.n.],2018:1-2.
[12] ZOU Jingui,WAN Ying,MENG Liyuan.A new stereo matching algorithm based on adaptive weight SAD algorithm and census algorithm[J].Bulletin of Surveying and Mapping,2018(11):11-15.(in Chinese)邹进贵,万荧,孟丽媛.一种基于自适应权重SAD与Census融合的匹配算法[J].测绘通报,2018(11):11-15.
[13] YU Fangjie,LUAN Xin,SONG Dalei,et al.An adaptive support-weight local stereo matching algorithm under variant illumination[J].Periodical of Ocean University of China,2014,44(4):108-111.(in Chinese)于方杰,栾新,宋大雷,等.一种光照变化条件下的自适用权重局部立体匹配算法[J].中国海洋大学学报(自然科学版),2014,44(4):108-111.
[14] GAO Jian,HUANG Xinhan,PENG Gang,et al.A feature detection method based on Harris corner and difference of Gaussian[J].Pattern Recognition and Artificial Intelligence,2008,21(2):171-176.(in Chinese)高健,黄心汉,彭刚,等.基于Harris角点和高斯差分的特征点提取算法[J].模式识别与人工智能,2008,21(2):171-176.
[15] FU Ying,ZENG Huanqiang,MA Lin,et al.Screen content image quality assessment using multi-scale difference of Gaussian[J].IEEE Transactions on Circuits and Systems for Video Technology,2018,28(9):2428-2432.
[16] KANADE T,OKUTOMI M.A stereo matching algorithm with an adaptive window:theory and experiment[J].IEEE Transactions on Pattern Analysis & Machine Intelligence,1994,16(9):920-932.
[17] ZHOU Xiuzhi,WEN Gongjian,WANG Runsheng.Fast stereo matching using adaptive window[J].Chinese Journal of Computers,2006,29(3):473-479.(in Chinese)周秀芝,文贡坚,王润生.自适应窗口快速立体匹配[J].计算机学报,2006,29(3):473-479.
[18] HUANG Chao,ZHAO Huazhi.Semi-global stereo matching with adaptive window based on grayscale value[J].Journal of Image and Graphics,2019,24(8):1381-1390.(in Chinese)黄超,赵华治.根据灰度值信息自适应窗口的半全局匹配[J].中国图象图形学报,2019,24(8):1381-1390.
[19] SEXTON W T,DULL C W,SZARO R C.Implementing ecosystem management:a framework for remotely sensed information at multiple scales[J].Landscape and Urban Planning,1998,40(3):173-184.
[20] ZHANG Dayu.The algorithm of WTA stereo matching based on the human detection method[J].Electronic Test,2013(5):66-68.
[21] CANNY J.A computational approach to edge detection[J].IEEE Transactions on Pattern Analysis and Machine Intelligence,1986,8(6):679-698.
[22] SCHARSTEIN D,SZELISKI R.A taxonomy and evaluation of dense two-frame stereo correspondence algorithms[J].International Journal of Computer Vision,2002,47(1/2/3):37-42.
[23] SCHARSTEIN D,SZELISKI R.High-accuracy stereo depth maps using structured light[C]//Proceedings of IEEE Computer Society Conference on Computer Vision and Pattern Recognition.Madison,USA:IEEE Press,2003:195-202.

选择文件类型/文献管理软件名称

选择包含的内容