SIFT优化算法及其在全景拼接图像配准中的应用
|
摘要
应用于全景拼接中的SIFT算法存在计算量较大、容易产生误匹配等不足,为提高全景拼接图像输出的效率,得到清晰的拼接图,提出一种优化的SIFT配准算法.优化算法首先采用基于拉普拉斯算子的图像边缘锐化处理,提取分块图像特征,再依据单元信息投影熵矢量欧式距离最小揣度进行特征匹配,最后利用改进的随机抽样一致性算法剔除错误匹配.实验证明,与原始SIFT配准算法相比,优化算法能够有效提高算法效率、减少错误匹配,取得了较好的匹配效果.
Aiming at overcoming the insufficients of SIFT registration algorithm such as large calculation,error matching and so on when applied in panoramic mosaic,an optimized SIFT registration algorithm was presented. For the optimized registration algorithm,Firstly,it sharpened the edge of image by bringing in the Laplacian and extracted characteristics of block images. Then matched feature points according to the minimum euclidean distance of entropy vector projection. Finally,it used improved Random Sample Consensus to eliminate the error matching. The optimized algorithm applicated in panoramic mosaic and experimental results showthat the optimized registration algorithm exceeded SIFT registration algorithm,it not only effectively improves the efficiency of algorithm,but also reduces the error matching,it achieves good matching effect.
Aiming at overcoming the insufficients of SIFT registration algorithm such as large calculation,error matching and so on when applied in panoramic mosaic,an optimized SIFT registration algorithm was presented. For the optimized registration algorithm,Firstly,it sharpened the edge of image by bringing in the Laplacian and extracted characteristics of block images. Then matched feature points according to the minimum euclidean distance of entropy vector projection. Finally,it used improved Random Sample Consensus to eliminate the error matching. The optimized algorithm applicated in panoramic mosaic and experimental results showthat the optimized registration algorithm exceeded SIFT registration algorithm,it not only effectively improves the efficiency of algorithm,but also reduces the error matching,it achieves good matching effect.
引文
[1]David G Lowe.Distinctive image features from scale invariant key points[J].International Journal of Computer Vision,2004,60(2):91-110.
[2]Kupfer B,Netanyahu N S,Shimshoni I.An efficient SIFT-based mode-seeking algorithm for sub-pixel registration of remotely sensed images[J].Geoscience and Remote Sensing Letters,IEEE,2014,12(2):379-383.
[3]Chen Meng-ting,Yan Dong-mei,Wang Gang.Algorithm of highresolution remote sensing image matching based on harris corner and SIFT descriptor[J].Chinese Journal of Image and Graphics,2012,17(11):1453-1459.
[4]Zheng Yue,Cheng Hong,Sun Wen-bang.Finding an optimal seamline through the shortest distance in the neighborhood[J].Chinese Journal of Image and Graphics,2014,19(2):227-233.
[5]Li Hai-chao,Hao Sheng-yong,Zhu Qi.Fast seamless mosaic algorithm for multiple remote sensing images[J].Infrared and Laser Engineering,2011,40(7):1381-1386.
[6]Ji Ying,Lu Yue-ming.Method based on SIFT for recognition of flip images[J].Application Research of Computers,2013,30(3):924-925,941.
[7]Yu Li-li,Dai Qing.Improved SIFT feature matching algorithm[J].Computer Engineering,2011,37(2):210-212.
[8]Wang Wan-tong,Han Zhi-gang,Liu Peng-fei.Multi-sensor image registration algorithm based on SIFT points and canny edge features matching[J].Computer Science,2011,38(7):287-289.
[9]Liu Yun-bo.Research on the key technology of image stitching[D].Chendu:University of Electronic Science and Technology,2013.
[10]Bao Paul.Image mosaics with wavelet domain seam-lines[C].IEEE Symposium on Computers and Informatics,2011:419-424.
[11]Zhang De-xiang,Zhang Jing-jing,Wu Xiao-pei,et al.Fusion of polarization image based on directionlets transform[J].Journal of Electronics and Information,2011,33(12):2795-2800.
[12]Su Xiu-qin,Liang Jin-feng.An improved target matching algorithm based on the unit of information entropy[J].Acta Photonica Sinica,2009,38(11):3040-3043.
[13]Fischler M A,Bolles R C.Random sample consensus:a paradigm for model fitting w ith applications to image analysis and automated acrtography[J].Communications of the ACM,1981,24(6):381-395.
[14]Lu Shan,Lei Ying-jie,Kong Wei-wei,et al.Fundamental matrix estimation based on probability analysis and sampling consensus[J].Control and Decision,2012,27(3):425-430.
[3]陈梦婷,闫冬梅,王刚.基于Harris角点和SIFT描述符的高分辨率遥感影像匹配算法[J].中国图象图形学报,2012,17(11):1453-1459.
[4]郑悦,程红,孙文邦.邻域最短距离法寻找最佳拼接缝[J].中国图象图形学报,2014,19(2):227-233.
[5]李海超,郝胜勇,朱琦.多片遥感图像的快速无缝拼接方法[J].红外与激光工程,2011,40(7):1381-1386.
[6]及莹,陆月明.一种基于SIFT算法的图像镜像变换识别方法[J].计算机应用研究,2013,30(3):924-925,941.
[7]于丽莉,戴青.一种改进的SIFT特征匹配算法[J].计算机工程,2011,37(2):210-212.
[8]王万同,韩志刚,刘鹏飞.基于SIFT点特征和Canny边缘特征匹配的多源遥感影像配准研究[J].计算机科学,2011,38(7):287-289.
[9]柳运波.全景图像拼接关键技术研究[D].成都:电子科技大学,2013.
[11]张德祥,张晶晶,吴小培,等.基于Directionlets变换的偏振图像融合[J].电子与信息学报,2011,33(12):2795-2800.
[12]苏秀琴,梁金峰.一种基于单元信息熵的目标匹配改进算法[J].光子学报,2009,38(11):3040-3043.
[14]鲁珊,雷英杰,孔韦韦,等.基于概率抽样一致性的基础矩阵估计算法[J].控制与决策,2012,27(3):425-430.
[2]Kupfer B,Netanyahu N S,Shimshoni I.An efficient SIFT-based mode-seeking algorithm for sub-pixel registration of remotely sensed images[J].Geoscience and Remote Sensing Letters,IEEE,2014,12(2):379-383.
[3]Chen Meng-ting,Yan Dong-mei,Wang Gang.Algorithm of highresolution remote sensing image matching based on harris corner and SIFT descriptor[J].Chinese Journal of Image and Graphics,2012,17(11):1453-1459.
[4]Zheng Yue,Cheng Hong,Sun Wen-bang.Finding an optimal seamline through the shortest distance in the neighborhood[J].Chinese Journal of Image and Graphics,2014,19(2):227-233.
[5]Li Hai-chao,Hao Sheng-yong,Zhu Qi.Fast seamless mosaic algorithm for multiple remote sensing images[J].Infrared and Laser Engineering,2011,40(7):1381-1386.
[6]Ji Ying,Lu Yue-ming.Method based on SIFT for recognition of flip images[J].Application Research of Computers,2013,30(3):924-925,941.
[7]Yu Li-li,Dai Qing.Improved SIFT feature matching algorithm[J].Computer Engineering,2011,37(2):210-212.
[8]Wang Wan-tong,Han Zhi-gang,Liu Peng-fei.Multi-sensor image registration algorithm based on SIFT points and canny edge features matching[J].Computer Science,2011,38(7):287-289.
[9]Liu Yun-bo.Research on the key technology of image stitching[D].Chendu:University of Electronic Science and Technology,2013.
[10]Bao Paul.Image mosaics with wavelet domain seam-lines[C].IEEE Symposium on Computers and Informatics,2011:419-424.
[11]Zhang De-xiang,Zhang Jing-jing,Wu Xiao-pei,et al.Fusion of polarization image based on directionlets transform[J].Journal of Electronics and Information,2011,33(12):2795-2800.
[12]Su Xiu-qin,Liang Jin-feng.An improved target matching algorithm based on the unit of information entropy[J].Acta Photonica Sinica,2009,38(11):3040-3043.
[13]Fischler M A,Bolles R C.Random sample consensus:a paradigm for model fitting w ith applications to image analysis and automated acrtography[J].Communications of the ACM,1981,24(6):381-395.
[14]Lu Shan,Lei Ying-jie,Kong Wei-wei,et al.Fundamental matrix estimation based on probability analysis and sampling consensus[J].Control and Decision,2012,27(3):425-430.
[3]陈梦婷,闫冬梅,王刚.基于Harris角点和SIFT描述符的高分辨率遥感影像匹配算法[J].中国图象图形学报,2012,17(11):1453-1459.
[4]郑悦,程红,孙文邦.邻域最短距离法寻找最佳拼接缝[J].中国图象图形学报,2014,19(2):227-233.
[5]李海超,郝胜勇,朱琦.多片遥感图像的快速无缝拼接方法[J].红外与激光工程,2011,40(7):1381-1386.
[6]及莹,陆月明.一种基于SIFT算法的图像镜像变换识别方法[J].计算机应用研究,2013,30(3):924-925,941.
[7]于丽莉,戴青.一种改进的SIFT特征匹配算法[J].计算机工程,2011,37(2):210-212.
[8]王万同,韩志刚,刘鹏飞.基于SIFT点特征和Canny边缘特征匹配的多源遥感影像配准研究[J].计算机科学,2011,38(7):287-289.
[9]柳运波.全景图像拼接关键技术研究[D].成都:电子科技大学,2013.
[11]张德祥,张晶晶,吴小培,等.基于Directionlets变换的偏振图像融合[J].电子与信息学报,2011,33(12):2795-2800.
[12]苏秀琴,梁金峰.一种基于单元信息熵的目标匹配改进算法[J].光子学报,2009,38(11):3040-3043.
[14]鲁珊,雷英杰,孔韦韦,等.基于概率抽样一致性的基础矩阵估计算法[J].控制与决策,2012,27(3):425-430.