视差约束的改进Hough变换多光谱影像匹配方法

doi:10.3969/j.issn.1001-2400.2015.06.017

J4 ›› 2015, Vol. 42 ›› Issue (6): 94-98.doi: 10.3969/j.issn.1001-2400.2015.06.017

视差约束的改进Hough变换多光谱影像匹配方法

赵宝玮1,2;相里斌1,2;黄旻2;吕群波2;张桂峰2

(1. 西安电子科技大学物理与光电工程学院，陕西西安 710071;
2. 中国科学院光电研究院，北京 100094)

收稿日期:2014-07-13 出版日期:2015-12-20 发布日期:2016-01-25
通讯作者: 赵宝玮
作者简介:赵宝玮(1983-), 男, 西安电子科技大学博士研究生, E-mail: zbw8200980@163.com．
基金资助:
国家杰出青年科学基金资助项目(61225024)

Matching method for multispectral images based on improved hough transformation with disparity constraint

ZHAO Baowei1,2;XIANGLI Bin1,2;HUANG Min2;LV Qunbo2;ZHANG Guifeng2

(1. School of Physics and Optoelectronic Engineering, Xidian Univ., Xi'an 710071, China;
2. Academy of Opto-electronics, Chinese Academy of Sciences, Beijing 100094, China)

Received:2014-07-13 Online:2015-12-20 Published:2016-01-25
Contact: ZHAO Baowei

摘要/Abstract

摘要：

针对机载滤光片型多光谱相机影像配准中同一目标在不同谱段表现的特征各不相同以及不同谱段图像存在较大灰度差异等问题，提出了视差约束的改进Hough变换多光谱影像匹配方法．该方法使用金子塔影像匹配策略，顶层金字塔影像通过尺度不变特征转换算子进行匹配来提供初始的视差约束条件，其他各层金字塔左右影像利用改进Hough变换影像方法进行匹配，匹配过程中利用上层匹配结果为当前层影像匹配提供视差约束条件，提高匹配速度及可靠性．通过对两幅1415×1415像素的机载多光谱飞行影像数据进行匹配，该方法仅耗时7.8s，配准精度能够达到065个像素．

关键词: 金字塔方法, 多光谱, 改进的Hough变换, 递进影像匹配

Abstract:

To overcome the problems of image matching caused by different charactors and the grey-level in every spectrum of the airborne filter multispectral imager for the same target， a matching method for multispectral images based on improved Hough transformation with disparity constraint is presented. Firstly, we get the N layer pyramid images of the two multispectral images for registration by down-sampling, and then calculate the transformation parameters of each layer between the left and right images progressively. In this progress, the transformation parameters of the two adjacent top Pyramid images can be calculated by using the feature points extracted by the SIFT operator, and the transformation parameters of the other adjacent Pyramid images can be calculated by using the transformation parameters of their upper images and the improved Hough transformation. At the same time, the attitude parameters and the geometric transform relation calculated from the upper layer between the images can be used to improve the matching speed and reliability. By processing two airborne flight multispectral images with 1415×1415 pixels, this approach only takes 78 secends, and its registration accuracy can reach 065 pixels.

Key words: pyramid method, multispectral, improved Hough transformation, progressive image matching

中图分类号:

TP751

赵宝玮;相里斌;黄旻;吕群波;张桂峰. 视差约束的改进Hough变换多光谱影像匹配方法[J]. J4, 2015, 42(6): 94-98.

ZHAO Baowei;XIANGLI Bin;HUANG Min;LV Qunbo;ZHANG Guifeng. Matching method for multispectral images based on improved hough transformation with disparity constraint[J]. J4, 2015, 42(6): 94-98.

参考文献

［1］徐晓轩, 林海波, 俞钢, 等. 小型线形可变滤光片分光的可见成像光谱仪及其特性研究［J］. 光谱学与光谱分析, 2002, 22(5): 714.
Xu Xiaoxuan, Lin Haibo, Yu Gang, et al. Reseatch of Image Spectrometer Using Linear Variable Interference Filter［J］. Spectroscopy and Spectral Analysis, 2002, 22(5): 714.
［2］ Darren T, Arko L, Zbyněk M, et al. Spatial Co-registration of Ultra-high Resolution Visible, Multispectral and Thermal Images Acquired with a Micro-UAV over Antarctic Moss Beds［J］. Remote Sensing, 2014, 6(5): 4003-4024.
［3］ Lee I H, Choi T S. Accurate Registration Using Adaptive Block Processing for Multispectral Images［J］. IEEE Transactions on Circuits And Systems for Video Technology, 2013, 23(9): 1491-1501.
［4］ Zitov B, Flusser J. Image Registration Methods: a Survey［J］. Image and Vision Computing, 2003, 21(11): 977-1000.
［5］ Wang F, Vemuri B C, Rao M. Cumulative Residual Entropy, a New Measure of Information ＆ Its Application to Image Alignment［J］. Computer Vision, 2003, 86(1): 112-114.
［6］钟灿. 非常规航摄影像定位方法及精度评定［D］. 武汉: 武汉大学, 2013.
［7］ Lowe D G. Distinctive Image Features from Scale-invariant Interest Points［J］. International Journal of Computer Vision, 2004, 60(2): 91-110.
［8］ Sajid S, Robert S. A Robust SIFT Descriptor for Multispectral Images［J］. IEEE Signal Processing Letters, 2014, 21(4): 400-403.
［9］ Li A L, Zhang B H. Registration Algorithm of Panchromatic and Multi-spectral Images Based on SIFT and RANSAC［C］//Symposium on Photonics and Optoelectronics. Piscataway: IEEE, 2012: 1-4.
［10］罗文超, 刘国栋, 杨海燕. SIFT和改进的RANSAC算法在图像配准中的应用［J］. 计算机工程与应用, 2013, 49 (15): 147-149.
Luo Wenchao, Liu Guodong, Yang Haiyan. Application of SIFT and Advanced RANSAC Algorithm on Image Registration［J］. Computer Engineering and Applications, 2013, 49(15): 147-149.
［11］ Goncalves H, Corte-Real L, Goncalves J A. Automatic Image Registration through Image Segmentation and SIFT［J］. IEEE Transactions on Geoscience and Remote Sensing, 2011, 49(7): 2589-2600.
［12］ Schneider M, Suri S, Reinartz P, et al. Matching of High-resolution Optical Data to a Shaded DEM［C］//International Geoscience and Remote Sensing Symposium. Piscataway: IEEE, 2009: 37-40.

[1]	刘帅;马文萍;杨淑媛;陈璞花. 结合分层字典学习和空谱信息的多光谱图像去噪[J]. 西安电子科技大学学报, 2017, 44(4): 174-180.
[2]	梁玮;曾平;郑海红;罗雪梅. 利用复合变换的多光谱图像压缩算法[J]. J4, 2015, 42(4): 33-40.
[3]	刘鑫;张建奇;王晓蕊;刘冬冬;张福娣. 一种多光谱热成像系统作用距离的预测方法[J]. J4, 2013, 40(2): 153-158.
[4]	侯彦宾;屈晓超;陈多芳;陈雪利. 一种新的多光谱光学参数等效方法[J]. J4, 2012, 39(4): 62-66+196.
[5]	罗雪梅;曾平;王莹;郑海红. 色外观匹配的多光谱图像再现[J]. J4, 2011, 38(4): 118-123.
[6]	王莹;曾平;何秀芳;罗雪梅. 采用查找表的超四色打印光谱色彩校正方法[J]. J4, 2010, 37(1): 41-48.
[7]	王柯俨1;吴成柯1;邓家先2;孔繁锵1;郭杰1. 基于经验数据分解的干涉多光谱图像压缩算法 [J]. J4, 2007, 34(6): 900-905.
[8]	肖江1;吴成柯1;李云松1;相里斌2;杨建峰2. 多光谱图像的3D EBCOT压缩编码算法[J]. J4, 2005, 32(4): 549-554.

视差约束的改进Hough变换多光谱影像匹配方法

Matching method for multispectral images based on improved hough transformation with disparity constraint

PDF (PC)

赞

可视化

被引次数

摘要/Abstract

引用本文

使用本文

参考文献

相关文章 8

Metrics

本文评价

推荐阅读 10