文章基本信息

标题：AN ADVANCED SCALE INVARIANT FEATURE TRANSFORM ALGORITHM FOR FACE RECOGNITION
本地全文：下载
作者：Mohammad Mohsen Ahmadinejad ; Elizabeth Sherly
期刊名称：Indian Journal of Computer Science and Engineering
印刷版ISSN：2231-3850
电子版ISSN：0976-5166
出版年度：2016
卷号：7
期号：3
页码：82-90
出版社：Engg Journals Publications
摘要：In computer vision, Scale-invariant feature transform (SIFT) algorithm is widely used todescribe and detect local features in images due to its excellent performance. But for face recognition, theimplementation of SIFT was complicated because of detecting false key-points in the face image due toirrelevant portions like hair style and other background details. This paper proposes an algorithm forface recognition to improve recognition accuracy by selecting relevant SIFT key-points only that byrejecting false key points. In the new proposed Haar-Cascade SIFT algorithm (HC-SIFT), the accuracyin face recognition has been increased from 52.6% to 75.1% from SIFT to HC- SIFT algorithm.
关键词：SIFT algorithm; Haar-Cascade; HC-SIFT; face recognition system; keypoint extraction