图像快速配准算法的改进

作者：谢永胜，余正生 日期：2010-03-26/span> 浏览：3734 查看PDF文档

图像快速配准算法的改进

谢永胜，余正生
（杭州电子科技大学 图形图像研究所，浙江 杭州 310018）

摘要：针对基于灰度的图像配准速度较慢的弊端，给出了一种改进的自适应阈值的快速淘汰非匹配点的配准算法，利用灰度误差和噪声相结合的优点，并且在此基础上引入多线程并行编程技术。实验结果表明，这种新型的配准改进算法在保证较高准确性的基础上大大地加快了配准速度，实时性能有较大提高。与先前提出的金字塔算法以及图像迹算法等一些算法相比，配准速度具有比较明显的改进。
关键词：图像配准；序贯相似算法；多线程编程；三次卷积插值法
中图分类号：TP391文献标识码：A文章编号：1001-4551(2010)02-0036-03

Improvement about fast image matching algorithm
XIE Yong-sheng, YU Zheng-sheng
(Graphics Institute,Hangzhou Dianzi University, Hangzhou 310018, China)
Abstract: Aiming at the drawback of slow registration based on gray-scale image, an improved fast adaptive threshold registration algorithm which could eliminated non-matching point was proposed, the privilege of the combination about gray-scale error and noise was used. Multi-threaded parallel programming technology was introduced. The experimental results show that, real-time performance has apparently improved by using the new improved algorithm for matching, and the match speed has been greatly accelerated with high accuracy. Compared with some of the previous algorithm, like the pyramid algorithm and image trace algorithm, the speed of registration has more obvious improvements.
Key words: image matching; sequential similarity detection algorithm; multi-threaded programming; cubic convolution interpolation
参考文献(References):
［1］HUANG X, PARAGIOS N, METAXAS D, Shape registrati-on in implicit spaces using information theory and free form deformations[J]. IEEE Trans.Pattern Anal.Mach,Intell,2006,28(8):1303-1318.
［2］LIU L, JIANG T, YANG J, et al. Fingerprint regist-ration by maximization of mutual information[J]. IEEE Trans.Image Process,2006,15(5):1100-1110.
［3］WANG L, GREENSPAN M, ELLIS R. Validation of bone segmentation and improved 3-d registration using contour coherency in ct data[J]. IEEE Trans.Med.Imaging,2006, 25(3):324-334.
［4］SCHWARZ D, KASPARED T, PROVAZNIK I, et al. A deformable registration method for automated morphometry of mri brain images in neuropsychi-atric research[J]. IEEE Trans.Med.Imaging,2007,26(4):452-461.
［5］BENTOUTOU Y, TABLE N, KPALMA K, et al. An auto-matic image registration for applications in remote sensing[J]. IEEE Trans. Geosci.Remote Sens,2005,43(9):2127-2137.
［6］BROWN L G. A survey of image registrtion techniques［J］. ACM Computing Surveys,1992,24(4):325-367.
［7］ZITOVA B, FLUSSER I. Image resistration methods a survey［J］. Image and Vision Computing,2003,21(11):977-1000.
［8］HUTTENLOCHER D P, KLANDERMAN G A, RUCKLIDGE W J. Comparing image using the distance［J］. Hausdo-rff.IEEE Transactions PAMI,1993,15(9):850-863.
［9］HONG Zheng-hua, ZHU Pei-ying. An improved SSDA app-Ied in target tracking［C］//Pattern Recognit-Ion,1988,9th IEEE International Conference,1988:767-769.
［10］HATABU A, MIYAZAKI T, KURODA I. Optimization of decision-timing for early ter-Mination of SSDA-based block matching［C］. Acoustics, Speech, and Signal Processing, Proce-Edings(ICASSP’03),2003 IEEE International Conference,2003.
［11］解凯，郭恒业，张田文.图像Mosaic技术综述［J］.电子学报，2004，32（4）：630-633.
［12］张立，周泽魁.基于特征融合的AVG小车图像数字识别方法[J].机电工程，2007，24(4)：4-6.
［13］刘小军，周越.基于轮廓特征的SAR图像自动配准［J］.计算机工程，2007，33（4）：176-178.
［14］刘文萍,吴立德.图像分割中阈值选择方法比较研究［J］.模式识别与人工智能,1997,10(3):271-277.
［15］章毓晋.图像处理和分析［M］.北京:清华大学出版社,1999.