摘要
首先采用稀疏点匹配算法提取待匹配的两幅图像的轮廓,构建轮廓高斯金字塔,然后逐层进行SIFT特征匹配。仿真实验结果表明,该方法提高了匹配速度和准确性。
First the sparse point matching algorithm is applied to extract the contour of two images to build outline Gaussian Pyramids,and then SIFT feature matching is used layer by layer.Simulation results indicate the method can improve both matching speed and accuracy.
出处
《长春工业大学学报》
CAS
2017年第1期58-61,共4页
Journal of Changchun University of Technology