Please wait a minute...
浙江大学学报(工学版)
浙江大学学报(工学版)
机械能源工程     
改进导向滤波器立体匹配算法
王志, 朱世强, 卜琰, 郭振民
1.浙江大学 流体动力与机电系统国家重点实验室,浙江 杭州 310027；
2.杭州自动化技术研究院有限公司,浙江 杭州 310030
Stereo matching algorithm using improved guided filtering
WANG Zhi, ZHU Shi qiang, BU Yan, GUO Zhen min
1. State Key Laboratory of Fluid Power and Mechatronic Systems, Zhejiang University, Hangzhou 310027, China;
2. Hangzhou Automation Technology Institute Co. Ltd, Hangzhou 310030, China
 全文: PDF(2230 KB)   HTML
摘要：

针对基于滤波器的立体匹配算法精度不高以及易受外界环境影响的问题,提出基于改进导向滤波器的立体匹配算法.在传统梯度向量中加入经过预处理后图像的梯度信息,结合Census变换计算匹配代价.构建自适应窗口,采用改进的导向滤波器聚合匹配代价|经过视差处理获得高精度的视差图,在低纹理区域能取得较好的匹配结果.实验结果表明,相比其他基于滤波器的立体匹配算法,该算法在Middlebury和KITTI平台上的测试结果具有更高的精度|对光照失真条件具有更好的鲁棒性,能应用于室外场合|计算复杂度与匹配窗口大小无关,具有较好的实时性.
Abstract:

An improved guided filter-based stereo matching algorithm was proposed to solve the problem of low accuracy and sensitivity to environmental change. The algorithm used the improved guiding filter to preprocess the raw images; the gradient of the preprocessed images was combined with the Census transform in the matching cost computation. Furthermore, an adaptive cross-based support window was constructed dependeding on the color similarity. The improved guided filter was adopted as the cost aggregation method. Accurate disparity maps were obtained after disparity refinement, which could achieve good performance in textureless regions. The experiments on the Middlebury and KITTI benchmark demonstrate that the proposed algorithm outperforms other filter-based methods, which has better robustness and can be used in outdoor applications. In addition, the computational complexity of the proposed method is independent on the window size, which has good real-time performance.
出版日期: 2016-12-08
:  TN 911.73  
基金资助:

国家自然科学基金资助项目(51521064)；杭州市创新链产业链重大科技创新资助项目(20132111A04)；杭州市重大科技创新资助项目(20142013A56).
通讯作者: 朱世强，男，教授，博导. ORCID: 0000-0003-1024-9769.     E-mail: sqzhu@sfp.zju.edu.cn
作者简介: 王志(1992—)，男，博士生，从事双目立体视觉研究. ORCID: 0000-0002-6236-6927. E-mail:11325067@zju.edu.cn
服务　　
把本文推荐给朋友
加入引用管理器
E-mail Alert
RSS
作者相关文章　　

引用本文:

王志, 朱世强, 卜琰, 郭振民. 改进导向滤波器立体匹配算法[J]. 浙江大学学报(工学版), 10.3785/j.issn.1008-973X.2016.12.003.

WANG Zhi, ZHU Shi qiang, BU Yan, GUO Zhen min. Stereo matching algorithm using improved guided filtering. JOURNAL OF ZHEJIANG UNIVERSITY (ENGINEERING SCIENCE), 10.3785/j.issn.1008-973X.2016.12.003.

链接本文:

http://www.zjujournals.com/eng/CN/10.3785/j.issn.1008-973X.2016.12.003    或    http://www.zjujournals.com/eng/CN/Y2016/V50/I12/2262

［1］ GURRIERI L E, DUBOIS E. Depth consistency and vertical disparities in stereoscopic panoramas ［J］. Journal of Electronic Imaging, 2014, 23(1): 114.
［2］ KUMAR S,KUMAR S,SUKAVANAM N,et al. Human visual system and segmentbased disparity estimation ［J］. AEUInternational Journal of Electronics and Communications, 2013, 67(5): 372-381.
［3］ SCHARSTEIN D, SZELISKI R. A taxonomy and evaluation of dense twoframe stereo correspondence algorithms ［J］. International Journal of Computer Vision, 2002, 47(1): 7-42.
［4］ 赖小波,朱世强.基于互相关信息的非参数变换立体匹配算法［J］.浙江大学学报：工学版,2011,45(9)：1636-1642.
LAI Xiaobo, ZHU Shiqiang. Mutual information based nonparametric transform stereo matching algorithm ［J］. Journal of Zhejiang University: Engineering Science, 2011, 45(9): 1636-1642.
［5］ 祝世平,杨柳.基于自适应分水岭的图割的立体匹配算法［J］.光学学报,2013,33(3)： 221-229.
ZHU Shiping, YANG Liu. Stereo matching algorithm with graph cuts based on adaptive watershed ［J］. Acta Optical Sinica, 2013, 33(3): 221-229.
［6］ TOMBARI F, MATTOCCIA S, DI S L, et al. Classification and evaluation of cost aggregation methods for stereo correspondence ［C］ ∥ IEEE Conference on Computer Vision and Pattern Recognition. Anchorage: IEEE, 2008: 18.
［7］ HOSNI A, BLEYER M, GELAUTZ M, et al. Local stereo matching using geodesic support weights ［C］ ∥ IEEE International Conference on Image Processing.Vienna: IEEE, 2009: 2093-2096.
［8］ VEKSLER O. Fast variable window for stereo correspondence using integral images ［C］ ∥ IEEE Conference on Computer Vision and Pattern Recognition. Madison: IEEE, 2003: 556-561.
［9］ ZHANG K, LU J, LAFRUIT G. Crossbased local stereo matching using orthogonal integral images ［J］. IEEE Transactions on Circuits and Systems for Video Technology, 2009, 19(7): 1073-1079.
［10］ YANG Y Y, WANG H B, LIU B. A new stereo matching algorithm based on adaptive window ［C］ ∥ Proceedings of IEEE International Conference on Systems and Informatics. Yantai: IEEE, 2012: 1815-1819.
［11］ YOON K J, KWEON I S. Adaptive supportweight approach for correspondence search ［J］. IEEE Transactions on Circuits and Systems for Video Technology, 2006, 28(4): 650-656.
［12］ ZHU S, CAO D H, WU Y B, et al. Edgeaware dynamic programming based cost aggregation for robust stereo matching ［J］. Journal of Electronic Imaging, 2015, 24(4): 17.
［13］ RHEMANN C, HOSNI A, BLEYER M, et al. Fast costvolume filtering for visual correspondence and beyond ［C］ ∥ IEEE Conference on Computer Vision and Pattern Recognition. Providence: IEEE, 2011: 3017-3024.
［14］ He K, SUN J, TANG X. Guided image filtering ［J］. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2013, 35(6), 1397-1409.
［15］ HIRSCHMULLER H, SCHARSTEIN D. Evaluation of stereo matching costs on images with radiometric differences ［J］. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2009, 31(9): 1582-1599.
［16］ MEI X, SUN X, ZHOU M C, et al. On building an accurate stereo matching system on graphics hardware ［C］ ∥ Proceedings of IEEE International Conference on Computer Vision Workshops. Barcelona: IEEE, 2011: 467-474.
［17］ MEI X, SUN X, DONG W, et al. segmenttree based cost aggregation for stereo matching ［C］ ∥ IEEE Conference on Computer Vision and Pattern Recognition. Portland: IEEE, 2013: 313-320.
［18］ HAM B, MIN D, OH C, et al. probabilitybased rendering for view synthesis ［J］. IEEE Transactions on Image Processing, 2014, 23(2): 870-884.
［19］ NGUYEN V D, NGUYEN D D, SANG J L, et al. Local density encoding for robust stereo matching ［J］. IEEE Transactions on Circuits and Systems for Video Technology, 2014, 24(12): 2049-2062.
［20］ HIRSCHMULLER H, et al. Stereo processing by semiglobal matching and mutual information ［J］. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2008, 30(2): 328-341.
［21］ GEIGER A, ROSER M, URTASUN R. Efficient largescale stereo matching ［C］ ∥ Asian Conference on Computer Vision. New Zealand: ［s. n.］, 2010: 25-38.
［22］ ZHANG K, FANG Y, MIN D, et al. Crossscale cost aggregation for stereo matching ［C］ ∥ IEEE Conference on Computer Vision and Pattern Recognition. Columbus: IEEE, 2014: 1590-1597.
[1] 朱株,刘济林. 基于马尔科夫随机场的三维激光雷达路面实时分割[J]. 浙江大学学报(工学版), 2015, 49(3): 464-469.
[2] 吴一全,张晓杰,吴诗婳,张生伟. 基于混沌PSO或分解的二维最小误差阈值分割[J]. J4, 2011, 45(7): 1198-1205.
[3] 毛锋, 张树有, 施岳定. 多分辨轮廓匹配的成型制件外形检测[J]. J4, 2011, 45(6): 1013-1020.
[4] 葛鹏,李奇, 冯华君,徐之海,陈跃庭. 双三次样条插值联合变换相关器亚像素探测技术[J]. J4, 2010, 44(11): 2198-2202.
[5] 谢强军, 侯迪波, 黄平捷, 张光新, 周泽魁. 基于半隐差分的单参数水平集快速分割[J]. J4, 2010, 44(8): 1496-1501.
[6] 毛锋, 张树有, 黄长林. 图像散布图和小波多分辨分析的模具异物检测[J]. J4, 2009, 43(10): 1749-1756.
[7] 凌波, 顾伟康, 杜歆. H264整帧丢失下的错误隐藏机制[J]. J4, 2009, 43(09): 1732-1738.