基于RFNA和改进LBD的镜像线特征匹配方法

doi:10.16180/j.cnki.issn1007-7820.2023.10.005

摘要/Abstract

摘要：

针对物体和镜像之间的匹配问题,引入RNFA(Relative Number of False Alarms)边缘链检测方法获取更丰富的线段。文中提出一种改进的LBD(Line Band Descriptor)算法用于构建局部不变特征描述符,通过比较局部不变特征描述符获得初始匹配对。采用全局投影角度的筛选方式,并通过拟合投影中线的方式剔除初始匹配对中误匹配项。在完成全局投影角度的选取和投影中线的拟合后,放宽对局部不变特征描述符阈值的筛选以获得更多的匹配对,提升召回率。图像集仿真实验结果表明,文中所提算法在纹理较弱区域能够更好地识别线段,且能够在保证原算法性能的基础上获得更多的匹配对,提高5%左右的正确匹配率,并达到90%以上的召回率。

关键词: 边缘链检测, RNFA, 局部不变特征描述符, 改进LBD, 线特征匹配, 镜像, 图像金字塔, 特征提取

Abstract:

In view of the matching problem between objects and mirrors in images, the RNFA(Relative Number of False Alarms)edge chain detection method is introduced to obtain richer line segments. An improved LBD(Line Band Descriptor) algorithm is proposed for constructing local invariant feature descriptors, and initial matching pairs are obtained by comparing local invariant feature descriptors. The screening of the global projection angle is adopted and the false matches in the initial matching pairs are eliminated fitting the projection center line. After the selection of the global projection angle and the fitting of the projection median are completed, the screening of the local invariant feature descriptor threshold is relaxed to obtain more matching pairs and improve the recall rate. The experimental results of image set simulation show that the designed algorithm can better identify line segments in the weaker texture regions and obtain more matching pairs on the basis of the guaranteed performance of the original algorithm, which can improve the correct matching rate by about 5% and achieve a recall rate of over 90%.

Key words: edge chain detection, RNFA, local invariant feature descriptor, improved LBD, line segment matching, mirror, image pyramid, feature extraction

中图分类号:

TP391.41

高于科,章伟,胡陟,江鹏伟. 基于RFNA和改进LBD的镜像线特征匹配方法[J]. 电子科技, 2023, 36(10): 32-38.

GAO Yuke,ZHANG Wei,HU Zhi,JIANG Pengwei. Line Segment Matching Based on RFNA and Improved LBD of Mirror Image[J]. Electronic Science and Technology, 2023, 36(10): 32-38.

图/表 10

图1

图2

图3

图4

图5

表1

图6

表2

图7

图8

参考文献 19

[1]	贾迪, 朱宁丹, 杨宁华, 等. 图像匹配方法研究综述[J]. 中国图象图形学报, 2019, 24(5):677-699.
	Jia Di, Zhu Ningdan, Yang Ninghua, et al. Image matching methods[J]. Journal of Image and Graphics, 2019, 24(5):677-699.
[2]	Campos C, Elvira R, Rodríguez J J G, et al. Orb-slam3:An accurate open-source library for visual,visual-inertial,and multimap slam[J]. IEEE Transactions on Robotics, 2021, 37(6):1874-1890. doi: 10.1109/TRO.2021.3075644
[3]	陈天泽, 李燕. 一种高性能SAR图像边缘点特征匹配方法[J]. 自动化学报, 2013, 39(12):2051-2063. doi: 10.3724/SP.J.1004.2013.02051
	Chen Tianze, Li Yan. A high performance edge point feature match method for SAR images[J]. Acta Automatica Sinica, 2013, 39(12):2051-2063. doi: 10.3724/SP.J.1004.2013.02051
[4]	Li Y, Wang F, Stevenson R, et al. Reliable line segment matching for multispectral images guided by intersection matches[J]. IEEE Transactions on Circuits and Systems for Video Technology, 2018, 29(10):2899-2912. doi: 10.1109/TCSVT.76
[5]	王招娣, 王贤立, 杨数强. 基于颜色和SURF特征的图像匹配算法[J]. 电子科技, 2017, 30(12):67-70.
	Wang Zhaodi, Wang Xianli, Yang Shuqiang. Image matching algorithm based on color and SURF feature[J]. Electronic Science and Technology, 2017, 30(12):67-70.
[6]	Wang L, Qian Y, Kong X. Line and point matching based on the maximum number of consecutive matching edge segment pairs for large viewpoint changing images[J]. Signal, Image and Video Processing, 2022, 16(1):11-18. doi: 10.1007/s11760-021-01959-6
[7]	Akinlar C, Topal C. EDLinezs:A real-time line segment detector with a false detection control[J]. Pattern Recognition Letters, 2011, 32(13):1633-1642. doi: 10.1016/j.patrec.2011.06.001
[8]	Von Gioi R G, Jakubowicz J, Morel J M, et al. LSD: A fast line segment detector with a false detection control[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2008, 32(4):722-732. doi: 10.1109/TPAMI.2008.300
[9]	Huang S Y, Shi Z Y, Zhong Y S. A simplified pipeline for line segment detection[C]. Shenyang: The Thirty-ninth Chinese Control Conference, 2020:1157-1162.
[10]	Lu X, Yao J, Li L, et al. Edge chain detection by applying Helmholtz principle on gradient magnitude map[C]. Cancun: The Twenty-third International Conference on Pattern Recognition, 2016:124-133.
[11]	Xing J, Wei Z, Zhang G. A line matching method based on multiple intensity ordering with uniformly spaced sampling[J]. Sensors, 2020, 20(6):1639-1650. doi: 10.3390/s20061639
[12]	Fan B, Wu F, Hu Z. Robust line matching through line-point invariants[J]. Pattern Recognition, 2012, 45(2):794-805. doi: 10.1016/j.patcog.2011.08.004
[13]	López J, Santos R, Fdez-Vidal X R, et al. Two-view line matching algorithm based on context and appearance in low-textured images[J]. Pattern Recognition, 2015, 48(7): 2164-2184. doi: 10.1016/j.patcog.2014.11.018
[14]	Zhang M, Yang Y, Jiang Q H, et al. A fast registration method based on line features[C]. Shenyang: The Thirty-ninth Chinese Control Conference, 2020:624-629.
[15]	Cheng Y, Jiang M, Zhu J, et al. Are we ready for unmanned surface vehicles in inland waterways?The USVinland multisensor dataset and benchmark[J]. IEEE Robotics and Automation Letters, 2021, 6(2):3964-3970. doi: 10.1109/LRA.2021.3067271
[16]	戚静瑜. 室内环境SLAM 中的镜面检测方法研究[D]. 北京: 北方工业大学, 2018:1-41.
	Qi Jingyu. Research on the method of mirror detection based on SLAM in indoor environment[D]. Beijing: North China University of Technology, 2018:1-41.
[17]	Zhang L, Koch R. An efficient and robust line segment matching approach based on LBD descriptor and pairwise geometric consistency[J]. Journal of Visual Communication and Image Representation, 2013, 24(7):794-805. doi: 10.1016/j.jvcir.2013.05.006
[18]	Mikolajczyk K, Schmid C. Scale and affine invariant interest point detectors[J]. International Journal of Computer Vision, 2004, 60(1):63-86. doi: 10.1023/B:VISI.0000027790.02288.f2
[19]	Wang Z, Wu F, Hu Z. MSLD:A robust descriptor for line matching[J]. Pattern Recognition, 2009, 42(5):941-953. doi: 10.1016/j.patcog.2008.08.035

标题	MP	EP	MP	EP	MP	EP	MP	EP	MRA/%
图像	图6(a)		图6(b)		图6(c)		图6(d)
LBD	102	2	25	2	100	2	14	1	95.22
改进LBD	106	1	26	0	102	1	13	0	99.52
图像	图6(e)		图6(f)		图6(g)		图6(h)
LBD	24	1	23	1	45	3	28	0	96.20
改进LBD	25	2	25	1	45	0	28	0	97.00
图像	图6(i)		图6(j)		图6(k)		图6(l)
LBD	41	5	50	8	56	16	17	1	84.34
改进LBD	42	5	54	9	59	11	19	1	86.88
图像	图6(m)		图6(n)		图6(o)		图6(p)
LBD	90	11	68	14	48	3	29	2	88.51
改进LBD	114	3	73	7	45	2	34	2	94.36

标题	CMP	TMP	CMP	TMP	CMP	TMP	CMP	TMP	RCR/%
图像	图6(a)		图6(b)		图6 (c)		图6 (d)
本文算法	110	115	32	34	104	109	16	18	94.93
图像	图6 (e)		图6 (f)		图6 (g)		图6 (h)
本文算法	24	24	29	30	52	55	33	35	95.83
图像	图6 (i)		图6 (j)		图6 (k)		图6 (l)
本文算法	41	42	49	51	55	62	23	28	91.80
图像	图6 (m)		图6 (n)		图6 (o)		图6 (p)
本文算法	133	142	74	80	50	50	33	37	93.85