摘要
为实现机车车底图像局部不变特征信号提取,比较了3种新兴的SIFT改进算法:GS-SIFT算法、RIT-SIFT算法和Harris-SIFT算法,对3种改进算法的准确性、实时性和高效性进行评价。应用基于SIFT的改进算法对火车机车车底实际图像进行处理,提取局部不变特征值,改进算法针对机车车底图像具有很好的适用性和准确高效性。
To extract local invariant feature signals of underbody images of locomotive,the Paper makes a comparison on three emerging improved SIFT algorithms：GS-SIFT Algorithm,RIT-SIFT Algorithm and Harris-SIFT Algorithm,making an evaluation on the accuracy,real-timeliness and efficiency of the three improved algorithms. It can process the actual underbody image by applying the improved algorithm that is based on SIFT,with the local invariant feature value extracted. The Improved algorithm is sound in applicability,accuracy and efficiency in view of underbody images of locomotive.
出处
《铁道技术监督》
2015年第2期40-45,共6页
Railway Quality Control
关键词
机车车底
图像分析
特征提取
局部不变
GS-SIFT
RIT-SIFT
Harris-SIFT
Underbody of Locomotive
Image Analysis
Feature Extraction
Local Invariant
GS-SIFT
RIT-SIFT
Harris-SIFT
