利用几何特征和小波变换实现图像的缩放匹配
Mosaic from Different-Sized Images Using Geometric Feature Mapping and Wavelet Transform
-
摘要: 通过对缩放图像几何信息的提取,利用B-spline函数的型值获取图像的缩放比例;同时,利用小波变换的性质简化了传统的模板匹配算法,最终实现了缩放图像的自动匹配实验结果表明,文中方法具有较快的运算速度和较好的匹配精度,在图像几何信息较为明显时具有较好的实用价值.Abstract: Scales of the images are matched through extracting their geometric information and adjusting the scale of control polygons of their B-spline contours. Meanwhile, wavelet transform is applied to simplify the computation of template matching algorithm. Mosaic of the images is built automatically. Experiments indicate that the algorithm is fast and accurate, it is most effective when the geometric information is distinct.
下载: