影像特征点检测提取算法研究
Research on image feature point detection and extraction algorithm
杨月1
作者信息
- 1. 辽宁省自然资源事务服务中心,辽宁 锦州 121003
- 折叠
摘要
影像特征点提取是影像分析与识别的重要基础,特征点检测是其中关键的环节.基于C++语言设计影像特征点自动提取程序,利用摄影测量里主流的Moravec算子、Harris算子以及Forstner算子进行影像特征点检测提取.通过试验数据对上述特征点检测算子在检测效率、抗噪性、适应性、准确性等方面进行性能评估,得出结论:Moravec算子和Forstner算子在检测效率方面上优势明显,Forstner算子的抗噪性能相对较好,上述3个算子在旋转变化、尺度变化和亮度变化上的适应性能力均较强.
Abstract
The image feature points extraction is an important basis for image analysis and recognition and the feature point detection is a key link.Based on C++language,an image feature point automatic extraction program is designed in this research to complete the image feature point detection and extraction process with such mainstreams as Moravec operator,Harris operator and Forstner operator in photogrammetry.Through the performance evaluation of the above feature point detection operators in terms of detection efficiency,noise resistance,adaptability,accuracy,etc.and the experimental data,it is concluded that both Moravec and Forstner algorithms have obvious advantages in detection efficiency,Forstner operator has relatively good noise resistance.Above mentioned three operators have stronger adaptiveness in rotation change,scale change,and brightness change.
关键词
影像特征点提取/检测算子/提取算法Key words
image feature points extraction/detection operator/extraction algorithm引用本文复制引用
出版年
2024
NETL
NSTL
万方数据