Mosaic Algorithm of Image Based on SURF and KLT Track
暂无分享,去创建一个
Aiming at the low precision and slow speed of registration,this paper propose an algorithm for automatic mosaic of image sequence.Using SURF vector is to judge if two images are overlapped and ascertain the relationship of the overlapped images.The detected features are regarded as moving objects,located by the KLT track algorithm,and accurate homography matrix between the two images is obtained.It proposes a color fusion method to achieve seamless splicing based on visual feature.Experimental results show that the precision and speed of registration are improved and the algorithm is robust.
[1] Wang Run-Sheng,et al. Fast RANSAC with Preview Model Parameters Evaluation , 2005 .
[2] Paul A. Viola,et al. Rapid object detection using a boosted cascade of simple features , 2001, Proceedings of the 2001 IEEE Computer Society Conference on Computer Vision and Pattern Recognition. CVPR 2001.
[3] G LoweDavid,et al. Distinctive Image Features from Scale-Invariant Keypoints , 2004 .
[4] Luc Van Gool,et al. Speeded-Up Robust Features (SURF) , 2008, Comput. Vis. Image Underst..