基于改进最佳缝合线的图像拼接方法
|
摘要
针对图像拼接中运动物体过于靠近缝合线造成重影和鬼影的现象,提出一种改进的最佳缝合线图像拼接方法。用SURF和ORB算法对特征点提取描述,采用PROSAC算法去除误匹配,利用提纯后的内点对待拼接图像进行RGB曝光差异矫正,用形态学操作和颜色饱和度差异改进能量函数,在最佳缝合线搜索准则中加入局部信息权重,结合融合带和加权平均融合的方法进行图像融合。实验结果表明,该方法能有效消除缝合线和鬼影,提高图像拼接质量。
To solve the phenomenon of ghosting when the moving object is too close to the best seam-line in the image mosaic,an improved image mosaic method based on the best seam-line was proposed.The feature points were extracted and described using SURF and ORB algorithms.PROSAC algorithm was used to eliminate false matching points.The exposure difference of the image was corrected with inliers.The cost function was improved by morphological operations and color saturation differences.The local information weight was added to the search criteria of the best seam-line.The fusion image was fused according to the fusion band and the weighted average fusion method.Experimental results show that the proposed method can effectively eliminate the seam-line and ghosting phenomenon and improve the quality of image mosaic.
To solve the phenomenon of ghosting when the moving object is too close to the best seam-line in the image mosaic,an improved image mosaic method based on the best seam-line was proposed.The feature points were extracted and described using SURF and ORB algorithms.PROSAC algorithm was used to eliminate false matching points.The exposure difference of the image was corrected with inliers.The cost function was improved by morphological operations and color saturation differences.The local information weight was added to the search criteria of the best seam-line.The fusion image was fused according to the fusion band and the weighted average fusion method.Experimental results show that the proposed method can effectively eliminate the seam-line and ghosting phenomenon and improve the quality of image mosaic.
引文
[1]ZHAO Xiaoqiang,YUE Zongda.Fast matching algorithm for image mosaic[J].Journal of Nanjing University of Science and Technology,2016,40(2):165-171(in Chinese).[赵小强,岳宗达.一种面向图像拼接的快速匹配算法[J].南京理工大学学报,2016,40(2):165-171.]
[2]ZOU Chengming,HOU Xiaobi,MA Jing.SIFT image stitching based on geometric image registration solution[J].Journal of Huazhong University of Science and Technology(Natural Science Edition),2016,44(4):32-36(in Chinese).[邹承明,侯小碧,马静.基于几何学图像配准的SIFT图像拼接算法[J].华中科技大学学报(自然科学版),2016,44(4):32-36.]
[3]JIANG Daihong,HUA Gang,WANG Yongxing.Research of automatic and quick stitching algorithm of mine monitoring image[J].Industry and Mine Automation,2015,41(4):78-82(in Chinese).[姜代红,华钢,王永星.矿井监控图像自动快速拼接算法研究[J].工矿自动化,2015,41(4):78-82.]
[4]ZHANG Yong,WANG Zhifeng,MA Wen.Research image mosaic algorithm based on improved SIFT feature matching[J].Microelectronics&Computer,2016,33(3):60-64(in Chinese).[张勇,王志峰,马文.基于改进SIFT特征点匹配的图像拼接算法研究[J].微电子学与计算机,2016,33(3):60-64.]
[5]YAN Zigeng,JIANG Jianguo,GUO Dan.Image matching based on SURF feature and Delaunay triangular meches[J].Acta Automatica Sinica,2014,40(6):1216-1222(in Chinese).[闫自庚,蒋建国,郭丹.基于SURF特征和Delaunay三角网格的图像匹配[J].自动化学报,2014,40(6):1216-1222.]
[6]Rublee E,Rabaud V,Konolige K,et al.ORB:An efficient alternative to SIFT or SURF[C]//Proceedings of IEEE International Conference on Computer Vision,2011:2564-2571.
[7]GU Yu,ZHOU Yang,REN Gang,et al.Image stitching by combining optimal seam and multi-resolution fusion[J].Journal of Image and Graphics,2017,22(6):842-851(in Chinese).[谷雨,周阳,任刚,等.结合最佳缝合线和多分辨融合的图像拼接[J].中国图象图形学报,2017,22(6):842-851.]
[8]DANG Jianwu,ZONG Yan,WANG Yangping.Research on image mosaic optimization algorithm based on SIFT feature detection[J].Application Research of Computers,2012,29(1):329-332(in Chinese).[党建武,宗岩,王阳萍.基于SIFT特征检测的图像拼接优化算法研究[J].计算机应用研究,2012,29(1):329-332.]
[9]SUN Peishuang,HAN Qirui.Improved optional stitching line applied to image stitching[J].Computer Engineering and Applications,2016,52(21):218-221(in Chinese).[孙培双,韩其睿.彩色图像最优拼接线的改进算法研究[J].计算机工程与应用,2016,52(21):218-221.]
[10]ZOU Lihui,CHEN Jie,ZHANG Juan,et al.Image mosaicing algorithm for dynamic scenes using multi-scaled PHOG feature and optimal seam[J].Pattern Recognition and Artificial Intelligence,2012,25(4):624-631(in Chinese).[邹丽辉,陈杰,张娟,等.一种多尺度PHOG特征和最优缝合线的运动场景图像拼接算法[J].模式识别与人工智能,2012,25(4):624-631.]
[11]MA Lili,CAO Chunmei,CHEN Jinguang,et al.Feature point matching algorithm based on RANSAC[J].Computer Engineering and Design,2016,37(7):1793-1797(in Chinese).[马丽丽,曹春梅,陈金广,等.基于RANSAC的特征点匹配算法[J].计算机工程与设计,2016,37(7):1793-1797.]
[12]LU Xiaojing,HUANG Xiangsheng.Fast calculation method of space transform model[J].Computer Science,2014,41(3):279-281(in Chinese).[路晓静,黄向生.一种快速的空间变换模型计算方法[J].计算机科学,2014,41(3):279-281.]
[13]QU Zhong,QIAO Gaoyuan,LIN Sipeng.Fast image stitching algorithm eliminates seam line and ghosting[J].Computer Science,2015,42(3):280-283(in Chinese).[瞿中,乔高元,林嗣鹏.一种消除图像拼接缝和鬼影的快速拼接算法[J].计算机科学,2015,42(3):280-283.]
[14]QIN Xujia,WANG Qi,WANG Huiling,et al.Image fusion method based on best seam-line for serial remote sensing images mosaic[J].Computer Science,2015,42(10):306-310(in Chinese).[秦绪佳,王琪,王慧玲,等.基于最佳缝合线的序列遥感图像拼接融合方法[J].计算机科学,2015,42(10):306-310.]
[2]ZOU Chengming,HOU Xiaobi,MA Jing.SIFT image stitching based on geometric image registration solution[J].Journal of Huazhong University of Science and Technology(Natural Science Edition),2016,44(4):32-36(in Chinese).[邹承明,侯小碧,马静.基于几何学图像配准的SIFT图像拼接算法[J].华中科技大学学报(自然科学版),2016,44(4):32-36.]
[3]JIANG Daihong,HUA Gang,WANG Yongxing.Research of automatic and quick stitching algorithm of mine monitoring image[J].Industry and Mine Automation,2015,41(4):78-82(in Chinese).[姜代红,华钢,王永星.矿井监控图像自动快速拼接算法研究[J].工矿自动化,2015,41(4):78-82.]
[4]ZHANG Yong,WANG Zhifeng,MA Wen.Research image mosaic algorithm based on improved SIFT feature matching[J].Microelectronics&Computer,2016,33(3):60-64(in Chinese).[张勇,王志峰,马文.基于改进SIFT特征点匹配的图像拼接算法研究[J].微电子学与计算机,2016,33(3):60-64.]
[5]YAN Zigeng,JIANG Jianguo,GUO Dan.Image matching based on SURF feature and Delaunay triangular meches[J].Acta Automatica Sinica,2014,40(6):1216-1222(in Chinese).[闫自庚,蒋建国,郭丹.基于SURF特征和Delaunay三角网格的图像匹配[J].自动化学报,2014,40(6):1216-1222.]
[6]Rublee E,Rabaud V,Konolige K,et al.ORB:An efficient alternative to SIFT or SURF[C]//Proceedings of IEEE International Conference on Computer Vision,2011:2564-2571.
[7]GU Yu,ZHOU Yang,REN Gang,et al.Image stitching by combining optimal seam and multi-resolution fusion[J].Journal of Image and Graphics,2017,22(6):842-851(in Chinese).[谷雨,周阳,任刚,等.结合最佳缝合线和多分辨融合的图像拼接[J].中国图象图形学报,2017,22(6):842-851.]
[8]DANG Jianwu,ZONG Yan,WANG Yangping.Research on image mosaic optimization algorithm based on SIFT feature detection[J].Application Research of Computers,2012,29(1):329-332(in Chinese).[党建武,宗岩,王阳萍.基于SIFT特征检测的图像拼接优化算法研究[J].计算机应用研究,2012,29(1):329-332.]
[9]SUN Peishuang,HAN Qirui.Improved optional stitching line applied to image stitching[J].Computer Engineering and Applications,2016,52(21):218-221(in Chinese).[孙培双,韩其睿.彩色图像最优拼接线的改进算法研究[J].计算机工程与应用,2016,52(21):218-221.]
[10]ZOU Lihui,CHEN Jie,ZHANG Juan,et al.Image mosaicing algorithm for dynamic scenes using multi-scaled PHOG feature and optimal seam[J].Pattern Recognition and Artificial Intelligence,2012,25(4):624-631(in Chinese).[邹丽辉,陈杰,张娟,等.一种多尺度PHOG特征和最优缝合线的运动场景图像拼接算法[J].模式识别与人工智能,2012,25(4):624-631.]
[11]MA Lili,CAO Chunmei,CHEN Jinguang,et al.Feature point matching algorithm based on RANSAC[J].Computer Engineering and Design,2016,37(7):1793-1797(in Chinese).[马丽丽,曹春梅,陈金广,等.基于RANSAC的特征点匹配算法[J].计算机工程与设计,2016,37(7):1793-1797.]
[12]LU Xiaojing,HUANG Xiangsheng.Fast calculation method of space transform model[J].Computer Science,2014,41(3):279-281(in Chinese).[路晓静,黄向生.一种快速的空间变换模型计算方法[J].计算机科学,2014,41(3):279-281.]
[13]QU Zhong,QIAO Gaoyuan,LIN Sipeng.Fast image stitching algorithm eliminates seam line and ghosting[J].Computer Science,2015,42(3):280-283(in Chinese).[瞿中,乔高元,林嗣鹏.一种消除图像拼接缝和鬼影的快速拼接算法[J].计算机科学,2015,42(3):280-283.]
[14]QIN Xujia,WANG Qi,WANG Huiling,et al.Image fusion method based on best seam-line for serial remote sensing images mosaic[J].Computer Science,2015,42(10):306-310(in Chinese).[秦绪佳,王琪,王慧玲,等.基于最佳缝合线的序列遥感图像拼接融合方法[J].计算机科学,2015,42(10):306-310.]