摘要
研究了图像配准中常用的几种特征点提取方法。基于图像配准对特征提取方法的特殊要求,提出了一组量化的特征提取方法评价因子,能用于合理评价各种方法的性能、优点与缺点,为这些方法的改进提供指导性意见和比较依据。通过大量的实验,利用评价因子比较和分析了上述的几种常用特征提取方法的优缺点、适用环境,并提出改进建议。
Several feature point extraction methods for image registration were studied. Some quantificational criterions for feature extraction methods evaluation were proposed based on the specific requirements of image registration processing. These criterions can be used to effectively evaluate the performance, advantages and disadvantages of various feature extraction methods and provide direction for further improvements. Through many experiments, advantages and disadvantages ＆ applicable environment of several feature point extraction methods mentioned were analyzed. The improved solutions are proposed.
出处
《光学技术》
EI
CAS
CSCD
北大核心
2007年第1期62-64,67,共4页
Optical Technique
基金
国防重点预研项目(51483040104BQ0101)
武器装备预研基金项目(514900201BQ0118)
关键词
图像配准
特征提取
特征匹配
image registration
feature extraction
feature matching
