基于种子点传播的快速立体匹配

胡汉平; 朱明

doi:10.3788/OPE.20152303.0887

您当前的位置：

首页 >

文章列表页 >

基于种子点传播的快速立体匹配

信息科学 | 更新时间：2020-08-13

- 基于种子点传播的快速立体匹配
- Fast stereo matching based on seed pixel propagation
- 光学精密工程 2015年23卷第3期页码：887-894
- 作者机构：
  
  1. 中国科学院长春光学精密机械与物理研究所,吉林长春,中国,130033
  2. 中国科学院大学北京,中国,100049
  3. 长春理工大学,吉林长春,130022
- 作者简介：
  
  [ "胡汉平(1980-),男,湖北襄阳人,博士研究生,工程师,2003年、2009年于长春理工大学分别获得学士、硕士学位,主要从事数字图像处理、计算机视觉等方面的研究。E-mail:custhhp@163.com" ]
  [ "朱明(1964-),男,江西南昌人,研究员,博士生导师,1985年于南京航空航天大学获学士学位,1991年于中国科学院长春光学精密机械与物理研究所获得硕士学位,主要从事视频图像处理和光电成像测量技术以及目标识别与电视跟踪技术方面的研究。E-mail:zhu-mingca@163.com" ]
- 基金信息：
  
  国家自然科学基金资助项目(No.61077156)()
- DOI：10.3788/OPE.20152303.0887
  中图分类号： TP391.4
- 收稿日期：2014-09-25，
  
  修回日期：2014-12-05，
  
  纸质出版日期：2015-03-25
- 稿件说明：
移动端阅览
胡汉平, 朱明,. 基于种子点传播的快速立体匹配[J]. 光学精密工程, 2015,23(3): 887-894

HU Han-ping, ZHU Ming,. Fast stereo matching based on seed pixel propagation[J]. Editorial Office of Optics and Precision Engineering, 2015,23(3): 887-894
胡汉平, 朱明,. 基于种子点传播的快速立体匹配[J]. 光学精密工程, 2015,23(3): 887-894 DOI： 10.3788/OPE.20152303.0887.

HU Han-ping, ZHU Ming,. Fast stereo matching based on seed pixel propagation[J]. Editorial Office of Optics and Precision Engineering, 2015,23(3): 887-894 DOI： 10.3788/OPE.20152303.0887.

摘要

针对计算机视觉中的对应点误匹配问题

提出了一种基于种子点传播的快速局部立体匹配算法来进一步提高匹配算法的运行效率。该算法首先利用Canny算子提取图像边缘

结合边缘信息构造动态匹配窗口

以克服固定窗口对匹配带来的不利影响; 然后利用AD-Census联合匹配代价在动态窗口上进行代价聚集

用WTA搜索策略得出初始视差图

对视差值进行筛选以确定种子点;随后利用像素间颜色差异将种子点的视差值传递给周围非种子点;最后采用区域投票和局部校正方式对视差值求精

进而获取精确的稠密视差图。实验结果表明

该算法可对Middlebury测试图生成高质量的视差图。与目前较新的局部立体匹配算法相比

其运行速度提高了1.8倍

满足了实际应用对速度和精度的要求

具有较高的实用价值。

Abstract

For the mismatching problem in computer vision

a fast local stereo matching algorithm based on seed pixel propagation was proposed to further improve the efficiency of matching algorithm. Firstly

the edge of an image was extracted using Canny operator in order to overcome the adverse influence of a fixed window on stereo matching by combining with the edge information to construct a dynamic matching window. Then

the AD-Census combined matching costs was used to realize the cost aggregation on the dynamic window

the initial disparity map was obtained by WTA search strategy and the disparity was filtrated to get seed pixels. Furthermore

the color difference between pixels was used to propagate the disparity of seed pixels to around unseeded pixels. Finally

the region voting and local rectification were adopted to refine the disparity and to get the accuracy dense disparity map. Experimental results show that the algorithm provides high-quality disparity map on Middlebury data set

the computing time is accelerated by 1.8 times as compared with that of newer local matching algorithms at present. It meets the demands of actual applications for accuracy and speeds.

关键词

Keywords

references

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.

卢阿丽, 唐振民. 利用自适应窗口实现不连续保护立体匹配 [J]. 光学精密工程, 2009, 17(9):2328-2334. LU A L, TANG ZH M, Discontinuity preserving stereo matching using variable window [J]. Opt. Precision Eng., 2009, 17(9):2328-2334. (in Chinese)

PHAM C, JEON J. Domain transformation-based efficient cost aggregation for local stereo matching [J]. IEEE, 2012, 23(7):1119-1130.

TATSUNORI T. Graph cut based continuous stereo matching using locally shared labels [C]. In CVPR, 2014.

周自维, 樊继壮, 李戈, 等. 支撑点扩展快速立体匹配方法的设计与应用 [J]. 光学精密工程, 2013, 21(1):207-215. ZHOU Z W, FAN J ZH, LI G, et al.. Design and application of fast matching method based on support point expansion [J]. Opt. Precision Eng., 2013, 21 (1):207-215. (in Chinese)

HIRSCHMULLER H, SCHARSTEIN D. Evaluation of stereo matching costs on images with radiometric differences [J]. IEEE, 2009, 31(9):1582-1599.

SUN X, MEI X, JIAO S, et al.. Stereo matching with reliable disparity propagation [C]. InProc:3DIMPVT, 2011:132-139.

HOSNI A, BLEYER M, GELAUTZ M. Secrets of adaptive support weight techniques for local stereo matching [C]. In CVIU, 2013, 117(6):620-632.

吴威, 许廷发, 王亚伟, 等. 高精度全景补偿电子稳像 [J]. 中国光学, 2013, 6(3):378-385. WU W, XU T F, WANG Y W, et al.. High precision digital image stabilization with full frame compensation [J]. Chinese Optics, 2013, 6(3):378-385. (in Chinese)

PAL C J, WEINMAN J J, TRAN L C, et al.. On learning conditional random fields for stereo:Exploring model structures and approximate inference [J]. International Journal of Computer Vision, 2012, 99(3):319-337.

吉淑娇, 朱明, 胡汉平. 基于特征点匹配的电子稳像技术 [J]. 中国光学, 2013, 6(6):841-849 JI SH J, ZHU M, HU H P. Aero-borne electronic image stabilization based on feature point matching [J]. Chinese Optics, 2013, 6(6):841-849. (in Chinese)

HU X Y, MORDOHAI P. A quantitative evaluation of confidence measures for stereo vision [J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2012, 34(11):2121-2133.

PFEIFFER D, GEHRIG S, SCHNEIDER N. Exploiting the power of stereo confidences [C]. IEEE Conference on Computer Vision and Pattern Recognition, 2013:297-304.

闫辉, 许廷发, 吴青青, 等. 多特征融合匹配的多目标跟踪 [J]. 中国光学, 2013, 6(2):163-170. YAN H, XU T F, WU Q Q, et al.. Multi-object tracking based on multi-feature joint matching [J]. Chinese Optics, 2013, 6(2):163-170.

KOWALCZUK J, PSOTA E, PEREZ L. Real-time stereo matching on CUDA using an iterative refinement method for adaptive support-weight correspondences [J]. IEEE Transactions on Circuits and Systems for Video Technology, 2013, 23(1):94-104.

SCHARSTEIN D, SZELISKI R. Middlebury stereo evaluation [EB/OL]. 2012, http://vision. middlebury. edu/stereo/eval/.

MIN D B, LU J B, DO M N, Joint Histogram-Based Cost Aggregation for Stereo Matching [J]. IEEE, 2013, 35(10):2539-2545.

LEE Z, JUANG J, NGUYEN T. Local disparity estimation with three-moded cross census and advanced support weight [J]. IEEE Transactions on Multimedia, 2013, 15(8):1855-1864.

ZHOU X Z, PIERRE BOULANGER. Radiometric invariant stereo matching based on relative gradients [C]. IEEE, 2012:2989-2992.

浏览量

255

下载量

CSCD

文章被引用时，请邮件提醒。

提交

工具集

关联资源

基于支撑点的立体匹配

空洞卷积和双边格网的立体匹配网络

面向小目标测量的通道注意力网络与系统设计

强噪声下自适应Canny算子边缘检测

利用平面约束的月球车立体匹配方法