基于双目视觉的改进特征立体匹配方法

doi:10.12263/DZXB.20200919

摘要/Abstract

摘要：

针对特征立体匹配方法只能得到稀疏视差以及弱纹理匹配率低以及视差精度不足导致视差连续处不平滑而呈阶梯状等问题，提出一种改进的特征立体匹配算法. 对预处理后的左右图像提取特征并进行特征匹配，再经过筛选获取准确的匹配点对；将得到的稀疏匹配点对作为种子点，依据视差连续性与极线约束准则建立一维搜索空间，利用积分图简化的快速零均值归一化互相关系数作为相似度量函数，通过双向匹配策略实现区域生长，大大提升匹配准确性的同时降低算法复杂度；通过亚像素拟合和加权中值滤波后处理提升视差精度，去除视差阶梯分层、噪声和条纹现象. Middlebury数据集实验结果表明，本算法得到了准确性更高且稠密的视差，提高了弱纹理区与深度不连续处的匹配效果以及整体视差精度，同时具有很强的鲁棒性，能抑制一定亮度差异和噪声的影响.

关键词: 双目视觉, 立体匹配, 稠密视差, 区域生长, 双向匹配, 视差精化

Abstract:

The feature-based stereo matching method can only get sparse disparity, and the low matching rate of weak texture areas and insufficient disparity accuracy lead to problems such as unsmooth continuity and stepped disparity. An improved features stereo matching algorithm is proposed. Extract features from the left and right images after preprocessing and perform feature matching, and then filter to obtain accurate matching point pairs; The obtained sparse matching point pairs are used as seed points, the search space is established according to the disparity continuity and the extreme line constraint criterion, and the fast zero-mean normalized cross-correlation simplified by integral graph is used as the similarity measurement function to achieve region growth through a two-way matching strategy, which greatly improves matching accuracy and reduces matching complexity; Sub-pixel fitting and weighted median filtering are used to improve the accuracy of disparity, remove disparity step layering, noise and streaks. The experimental results of the Middlebury data set show that this algorithm obtains a highly accurate and dense disparity, improves the matching effect of weak texture areas and depth discontinuities, and the accuracy of disparity. At the same time, it has strong robustness and can suppress the influence of certain brightness differences and noise.

Key words: binocular vision, stereo matching, dense disparity, regional growth, two-way matching, disparity refinement

中图分类号:

TP391

王笛, 胡辽林. 基于双目视觉的改进特征立体匹配方法[J]. 电子学报, 2022, 50(1): 157-166.

WANG Di, HU Liao-lin. Improved Feature Stereo Matching Method Based on Binocular Vision[J]. Acta Electronica Sinica, 2022, 50(1): 157-166.

图/表 15

参考文献 20

1	WANG X L, YANG L, WANG L R, et al. Characteristic point match algorithm based on the SURF in binocular stereo vision[C]//2012 Fifth International Conference on Intelligent Networks and Intelligent Systems. Piscataway NJ,USA: IEEE, 2012: 302-305.
2	YIN Z, XIONG J. Stereovision measurement of layer geometry in wire and arc additive manufacturing with various stereo matching algorithms[J]. Journal of Manufacturing Processes, 2020, 56: 428-438.
3	SU C, TAN G, LUO Y. Research on stereo matching technology based on binocular vision[J]. Open Access Library Journal, 2019, 6(9): 1-10.
4	CHENG J. Three-dimensional reconstruction of binocular stereo vision based on improved SURF algorithm and KD-Tree matching algorithm[J]. Journal of Digital Information Management, 2015, 13(6):462-466.
5	WANG N, PENG B B, ZHU D, et al. Stereo matching for binocular underwater images using SURF operator and epipolar restriction[C]//Life System Modeling and Simulation. Berlin Heidelberg,Ger: Springer, 2014: 378-386.
6	王阳萍, 秦安娜, 郝旗, 等. 结合加速鲁棒特征的遥感影像半全局立体匹配[J]. 光学学报, 2020, 40(16): 163-171.
	WANG Y P, QIN A N, HAO Q, et al. Semi-Global stereo matching of remote sensing images combined with speeded up robust features[J]. Acta Optica Sinica,2020, 40(16):163-171. (in Chinese)
7	FENG H, ZHANG G, HU B, et al. Noise-resistant matching algorithm integrating regional information for low-light stereo vision[J]. Journal of Electronic Imaging, 2019, 28(1): 013050.1-013050.10.
8	HUI J, YANG Y, HUI Y, et al. Research on identify matching of object and location algorithm based on binocular vision[J]. Journal of Computational and Theoretical Nanoscience, 2016, 13(3): 2006-2013.
9	肖进胜, 田红, 邹文涛, 等. 基于斜平面平滑优化的半全局立体匹配[J]. 电子学报, 2018, 46(8): 1835-1841.
	XIAO J S, TIAN H, ZOU W T, et al. Semi-global matching algorithm based on improved slanted plane smoothing[J]. Acta Electronica Sinica,2018, 46(8): 1835-1841. (in Chinese)
10	GUO S, YANG X. Fast recognition algorithm for static traffic sign information[J]. Open Physics, 2018, 16(1): 1149-1156.
11	MENG Q, LU X, ZHANG B, et al. Research on the ROI registration algorithm of the cardiac CT image time series[J]. Biomedical Signal Processing and Control, 2018, 40: 71-82.
12	LIN C, LI Y, XU G, et al. Optimizing ZNCC calculation in binocular stereo matching[J]. Signal Processing: Image Communication, 2017, 52: 64-73.
13	MEN Y, ZHANG G, MEN C, et al. A sub-pixel disparity refinement algorithm based on lagrange interpolation[J]. Chinese Journal of Electronics, 2017, 26(4): 784-789.
14	ZHANG Q, XU L, JIA J. 100+ times faster weighted median filter(WMF)[C]//2014 IEEE Conference on Computer Vision and Pattern Recognition. Columbus OH, USA: IEEE, 2014: 2830-2837.
15	毛昕蓉, 王楠. 基于ADCensus的改进双目立体匹配算法[J].电子产品世界, 2020, 27(4): 62-64, 87.
	MAO X R, WANG N. Improved binocular stereo matching algorithm based on ADCensus[J]. Electronics World, 2020, 27(4): 62-64+87.
16	ALLABERDIEV S, YAKHYOEV S, FATKHULLAYEV R, et al. Speeded-up robust feature matching algorithm based on image improvement technology[J]. Journal of Computer and Communications, 2019, 7(12): 1-10.
17	欧阳鑫玉, 张娟娟, 赵楠楠, 等. 基于NCC的改进立体匹配算法[J]. 微型机与应用, 2015, 34(3): 54-57.
	OUYANG X Y, ZHANG J J, ZHAO N N, et al. Improved stereo matching algorithm based on NCC[J]. Microcomputer & Its Applications, 2015, 34(3): 54-57. (in Chinese)
18	HUANG S, BI Y, XU X. Research and implementation of binocular stereo matching algorithms[J]. Journal of Ludong University(Natural Science Edition), 2018, 34(1): 25-30.
19	LIU Y, HUANG W, WANG Y. A new stereo matching method for RAW image data based on improved SGBM[C]// Three-Dimensional Image Acquisition and Display Technology and Applications. Bellingham,USA: SPIE, 2018: 108450W.
20	ZHU S, GAO R, LI Z. Stereo matching algorithm with guided filter and modified dynamic programming[J]. Multimedia Tools and Applications, 2017, 76(1): 199-216.

Image	Number of FAST feature point before pretreatment		Number of FAST feature point after pretreatment		Number of matching point pairs before/after Eliminate false matches
Image	Left	Right	Left	Right
Venus	536	526	627	637	78	74
Bull	221	223	384	340	40	40
Cones	515	519	640	587	54	53
Teddy	349	380	383	403	43	41

Image	Number of FAST feature point before pretreatment		Number of FAST feature point after pretreatment		Number of matching point pairs before/after Eliminate false matches
Image	Left	Right	Left	Right
Venus	536	526	627	637	78	74
Bull	221	223	384	340	40	40
Cones	515	519	640	587	54	53
Teddy	349	380	383	403	43	41

Image	Objective index	NCC	BM	SGBM	DP	Proposed algorithm
Venus	Error rate	0.179	0.219	0.056	0.275	0.020
Bull	Error rate	0.099	0.166	0.050	0.179	0.014
Cones	Error rate	0.236	0.324	0.193	0.368	0.131
Teddy	Error rate	0.286	0.369	0.232	0.452	0.148

Image	Objective index	NCC	BM	SGBM	DP	Proposed algorithm
Venus	Error rate	0.179	0.219	0.056	0.275	0.020
Bull	Error rate	0.099	0.166	0.050	0.179	0.014
Cones	Error rate	0.236	0.324	0.193	0.368	0.131
Teddy	Error rate	0.286	0.369	0.232	0.452	0.148

[1]	吴俊劼, 陈震, 张聪炫, 江少锋, 尚璇. 基于特征级联卷积网络的双目立体匹配[J]. 电子学报, 2021, 49(4): 690-695.
[2]	陈震, 马龙, 张聪炫, 黎明, 吴俊劼, 江少锋. 基于语义分割的双目场景流估计[J]. 电子学报, 2020, 48(4): 631-636.
[3]	马宁, 门宇博, 门朝光, 李香. 基于扩展相位相关的小基高比立体匹配方法[J]. 电子学报, 2017, 45(8): 1827-1835.
[4]	姜慧研, 冯锐杰. 基于改进的变分水平集和区域生长的图像分割方法的研究[J]. 电子学报, 2012, 40(8): 1659-1664.
[5]	赵士伟;卓力;王素玉;沈兰荪. 基于数据挖掘的图像压缩域肤色检测算法[J]. 电子学报, 2010, 38(3): 605-610.
[6]	郭大波;卢朝阳;焦卫东;崔玲玲. 遮挡检测/立体匹配中的分段动态规划法[J]. 电子学报, 2009, 37(7): 1516-1521.
[7]	伍春洪;杨扬;游福成. 一种基于Integral Imaging和多基线立体匹配算法的深度测量方法[J]. 电子学报, 2006, 34(6): 1090-1095.
[8]	邓毅;林学. 一种新的快速立体视觉导航算法[J]. 电子学报, 2006, 34(11): 2090-2093.
[9]	徐杰;施鹏飞. 基于相位一致与区域生长的自然彩色图像分割[J]. 电子学报, 2004, 32(7): 1203-1205.
[10]	管业鹏;童林夙;陈娜. 基于双目立体视觉的偏转线圈测量方法研究[J]. 电子学报, 2003, 31(9): 1382-1385.