摘要
图像拼接是计算机视觉领域的一个重要分支,也是现在的研究热点。研究了用于图像拼接的特征提取算法,该算法提取的特征能对旋转、尺度缩放、仿射变换、视角变化、光照变化、噪声等图像变化因素保持一定的不变性,并在处理速度上与传统的SIFT算法相比有大幅提高,具有较强的实时性和鲁棒性,是一种较好的特征匹配算法。
Image stitch is an important branch of computer vision, which is a focus these years. An arithmetic of feature extraction which used in image stitch is researched. The features extracted by this method are shown to be invariant to the image rotation and scale and robust across a substantial range of aft＇me distortion, addition of noise, and change in illumination. Besides above merits, this method is more quickly than traditional SIFT. That is an ideal method of feature matching.
出处
《计算机工程与设计》
CSCD
北大核心
2009年第2期440-442,共3页
Computer Engineering and Design
基金
江苏省"六大人才高峰"基金项目(07-E-024)
关键词
小波变换
SIFT
特征提取
图像拼接
混合编程
wavelet transform
SIFT
feature extraction
image stitch
hybrid-programming
