基于融合代价和优化引导滤波的立体匹配算法

doi:10.19678/j.issn.1000-3428.0064116

摘要/Abstract

摘要： 现有局部立体匹配算法与全局立体匹配算法相比，计算量更小、速度更快，能达到实时匹配的要求，但存在误匹配率较高、视差结果精度较低等问题。提出一种新的局部立体匹配算法，依据像素梯度信息、像素梯度的平均值及标准差计算多指标梯度代价，使算法对图像局部结构的表达更加全面。根据像素的灰度差异程度划分等级，提出7等级编码的精细化Census变换匹配代价，以有效识别图像信息之间的细微差异，弥补Census变换在相似纹理区域精度较低的不足。将多指标梯度代价和精细化Census变换匹配代价按一定权重进行融合，形成初始匹配代价，从而充分发挥两种代价的优越性。在代价聚合阶段，对引导滤波的线性系数进行自适应优化，解决固定参数引起图像局部过于平滑或平滑不足的问题，并使用优化后的引导滤波模型进行代价聚合，改善代价聚合的效果。使用胜者为王算法计算视差获得初始视差图，最后通过左右一致性检测和加权中值滤波进行视差优化，得到更为理想的视差图。在Middlebury V3立体评估平台上的实验结果表明，所提算法在非遮挡区域的加权平均误匹配率为15.7%，与Cens5、IGF、ISM等算法相比具有较高的精度。

关键词: 图像处理, 立体匹配, 梯度, Census变换, 引导滤波, 自适应优化

Abstract: The existing local stereo matching algorithm requires fewer computation and has a higher speed than global stereo matching, making it suitable for real-time matching.However, it has issues such as high rates of matching errors and poor accuracy of disparity results.A new local stereo matching algorithm is proposed.The multi-index gradient cost is calculated according to the pixel gradient information, the average, and standard deviation of the gradient.This allows the algorithm to better capture the local structure of the image.The proposed matching cost using 7-level coding of the refined Census transform can help identify small differences between images, which can fix the low accuracy of the Census transform in similar texture regions.Next, the multi-index gradient cost and refined Census transform matching cost are combined according to a certain weight to form the initial matching cost, thereby leveraging the advantages of the two costs.In the cost aggregation stage, the linear coefficients of the guided filtering are adaptively optimized to solve the problem of excessive or insufficient image smoothing caused by fixed parameters.The optimized guided filtering model is used for cost aggregation to improve the effect of cost aggregation.Then, the Winner-Take-All(WTA) algorithm calculates the disparity to obtain the initial disparity map.Finally, an ideal disparity map is obtained through disparity optimization through left-right consistency check and weighted median filtering. Experimental results on the Middlebury V3 stereo evaluation platform demonstrate that the proposed algorithm has a high accuracy compared with Cens5、IGF、ISM and other algorithm, with a weighted average percentage of bad matching pixels in the non-occlusion regions of 15.7%.

Key words: image processing, stereo matching, gradient, Census transformation, guided filtering, adaptive optimization

中图分类号:

TP391.41

余嘉昕, 王春媛, 韩华, 高燕. 基于融合代价和优化引导滤波的立体匹配算法[J]. 计算机工程, 2023, 49(3): 257-262,270.

YU Jiaxin, WANG Chunyuan, HAN Hua, GAO Yan. Stereo Matching Algorithm Based on Fusion Cost and Optimized Guided Filtering[J]. Computer Engineering, 2023, 49(3): 257-262,270.

http://www.ecice06.com/CN/Y2023/V49/I3/257

图/表 8

20230314190452

20230314190455

20230314190501

20230314190506

20230314190509

20230314190512

20230314190515

20230314190518

参考文献

[1] 陈炎, 杨丽丽, 王振鹏.双目视觉的匹配算法综述[J].图学学报, 2020, 41(5):702-708. CHEN Y, YANG L L, WANG Z P.Literature survey on stereo vision matching algorithms[J].Journal of Graphics, 2020, 41(5):702-708.(in Chinese)
[2] 杨冬冬, 张晓林, 李嘉茂.基于局部与全局优化的双目视觉里程计算法[J].计算机工程, 2018, 44(1):1-8. YANG D D, ZHANG X L, LI J M.Binocular visual odometry algorithm based on local and global optimization[J].Computer Engineering, 2018, 44(1):1-8.(in Chinese)
[3] 罗昌平, 夏海波, 赵琼.基于运动视差的三维重建技术在自导引小车中的应用[J].图学学报, 2017, 38(4):615-622. LUO C P, XIA H B, ZHAO Q.The application of three-dimensional reconstruction technique based on motion parallax in AGV[J].Journal of Graphics, 2017, 38(4):615-622.(in Chinese)
[4] 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.
[5] FELZENSZWALB P F, HUTTENLOCHER D R.Efficient belief propagation for early vision[C]//Proceedings of IEEE Computer Society Conference on Computer Vision and Pattern Recognition.Washington D.C., USA:IEEE Press, 2004:261-268.
[6] KIM J C, LEE K M, CHOI B T, et al.A dense stereo matching using two-pass dynamic programming with generalized ground control points[C]//Proceedings of IEEE Computer Society Conference on Computer Vision and Pattern Recognition.Washington D.C., USA:IEEE Press, 2005:1075-1082.
[7] BLEYER M, GELAUTZ M.Graph-cut-based stereo matching using image segmentation with symmetrical treatment of occlusions[J].Signal Processing:Image Communication, 2007, 22(2):127-143.
[8] OKUTOMI M, KANADE T.A multiple-baseline stereo[J].IEEE Transactions on Pattern Analysis and Machine Intelligence, 1993, 15(4):353-363.
[9] 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.Washington D.C., USA:IEEE Press, 2012:467-474.
[10] 王云峰, 吴炜, 余小亮, 等.基于自适应权重AD-Census变换的双目立体匹配[J].工程科学与技术, 2018, 50(4):153-160. WANG Y F, WU W, YU X L, et al.A stereo matching system with the adaptive weight AD-Census[J].Advanced Engineering Sciences, 2018, 50(4):153-160.(in Chinese)
[11] ZHAN Y L, GU Y Z, HUANG K, et al.Accurate image-guided stereo matching with efficient matching cost and disparity refinement[J].IEEE Transactions on Circuits and Systems for Video Technology, 2016, 26(9):1632-1645.
[12] ZHANG K, LU J B, 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.
[13] HOSNI A, RHEMANN C, BLEYER M, et al.Fast cost-volume filtering for visual correspondence and beyond[J].IEEE Transactions on Pattern Analysis and Machine Intelligence, 2013, 35(2):504-511.
[14] 季顺平, 罗冲, 刘瑾.基于深度学习的立体影像密集匹配方法综述[J].武汉大学学报(信息科学版), 2021, 46(2):193-202. JI S P, LUO C, LIU J.A review of dense stereo image matching methods based on deep learning[J].Geomatics and Information Science of Wuhan University, 2021, 46(2):193-202.(in Chinese)
[15] KENDALL A, MARTIROSYAN H, DASGUPTA S, et al.End-to-end learning of geometry and context for deep stereo regression[C]//Proceedings of IEEE International Conference on Computer Vision.Washington D.C., USA:IEEE Press, 2017:66-75.
[16] CHANG J R, CHEN Y S.Pyramid stereo matching network[C]//Proceedings of IEEE/CVF Conference on Computer Vision and Pattern Recognition.Washington D.C., USA:IEEE Press, 2018:5410-5418.
[17] ZHANG F H, PRISACARIU V, YANG R G, et al.GA-Net:guided aggregation net for end-to-end stereo matching[C]//Proceedings of IEEE/CVF Conference on Computer Vision and Pattern Recognition.Washington D.C., USA:IEEE Press, 2020:185-194.
[18] XU H F, ZHANG J Y.AANet:adaptive aggregation network for efficient stereo matching[C]//Proceedings of IEEE/CVF Conference on Computer Vision and Pattern Recognition.Washington D.C., USA:IEEE Press, 2020:1956-1965.
[19] HE K M, SUN J, TANG X O.Guided image filtering[J].IEEE Transactions on Pattern Analysis and Machine Intelligence, 2013, 35(6):1397-1409.
[20] SCHARSTEIN D, HIRSCHMÜLLER H, KITAJIMA Y, et al.High-resolution stereo datasets with subpixel-accurate ground truth[M]//JIANG X Y, HORNEGGER J, KOCH R.Lecture Notes in Computer Science.Berlin, Germany:Springer, 2014:31-42.
[21] SCHARSTEIN D, HIRSCHMÜLLER H.Middlebury stereo vision page[EB/OL].[2021-08-26].https://vision.middlebury.edu/stereo.
[22] HIRSCHMÜLLER H, INNOCENT P R, GARIBALDI J.Real-time correlation-based stereo vision with reduced border errors[J].International Journal of Computer Vision, 2002, 47(1):229-246.
[23] HAMZAH R A, IBRAHIM H, ABU HASSAN A H.Stereo matching algorithm based on per pixel difference adjustment, iterative guided filter and graph segmentation[J].Journal of Visual Communication and Image Representation, 2017, 42(1):145-160.
[24] HAMZAH R A, IBRAHIM H, ABU HASSAN A H.Stereo matching algorithm for 3D surface reconstruction based on triangulation principle[C]//Proceedings of the 1st International Conference on Information Technology, Information Systems and Electrical Engineering.Washington D.C., USA:IEEE Press, 2017:119-124.
[25] 孔令寅, 朱江平, 应三丛.基于引导图像和自适应支持域的立体匹配[J].光学学报, 2020, 40(9):86-98. KONG L Y, ZHU J P, YING S C.Stereo matching based on guidance image and adaptive support region[J].Acta Optica Sinica, 2020, 40(9):86-98.(in Chinese)

选择文件类型/文献管理软件名称

选择包含的内容