基于改进加权融合算法的运动场景图像拼接

在图像匹配、融合过程中,由于图像中运动物体的存在,使得融合后的图像可能出现鬼影。鬼影消除一直是图像拼接技术的重点和难点。针对这一问题,现提出一种基于SIFT和改进加权融合的运动场景图像拼接方法。首先,采用对尺度具有鲁棒性的SIFT算法进行特征点的提取与匹配,使用RANSAC算法求出图像间的变换矩阵H,最后利用文中提出的改进加权融合算法完成图像的无缝拼接。整个过程完全自动地实现运动场景的图像拼接,消除了运动鬼影。实验结果验证了该方法的有效性。

Image mosaics of dynamic scene based on SIFT and an improved smoothing algorithm

In the process of image registration and image fusion, according to the existence of the moving object in the image, the fused images may appear ghosting. The elimination of ghosting is always the key and difficulty of image stitching technology. To solve this problem, this paper presents a method for image mosaics of dynamic scene based on SIFT and an improved smoothing algorithm. Firstly, a scale- invariant feature extracting algorithm SIFT is used to for feature extraction and matching. The transforming matrix H is computed with RANSAC algorithm. And finally image mosaic is completed with the improved smoothing algorithm. The method implements automatically and eliminates ghosting. The experimental results show that, this method is stable and effective.