SIFT特征提取的指静脉认证技术研究
首发时间:2018-05-24
摘要:随着信息安全对人们来说越来越重要,生物识别技术基于人类的生物特征或者行为来识别,本身就具有可靠性。关于生物识别技术主要有虹膜,指纹,静脉等,而指静脉由于其易于获取,成本不高,具有一定的研究价值。SIFT算法由于其抗干扰,抗伸缩,抗旋转等特点,被广泛应用于图片的识别。但是由于指静脉由其信息量较少,采集的图片分辨率低,干扰较多,从而比手背静脉,指纹等难于识别。而其优点在于方便采集等。本文就遇到的问题分别对指静脉进行分割,二值化,平移进行校准,同时压缩特征向量减少其维数,同时取得了较好的识别率
For information in English, please click here
Research on finger vein authentication technology based on SIFT feature extraction
Abstract:As information security becomes more and more important to people, biometrics technology is based on human biometrics or behavior to identify itself with reliability. The biometric technologies mainly include irises, fingerprints, veins, etc. The finger veins have certain research value due to their easy access and low cost. SIFT algorithm is widely used in image recognition due to its anti-jamming, anti-scaling and anti-rotation features. However, due to the small amount of information in the finger veins, the resolution of the collected images is low, and the interference is much more difficult to identify than the hand veins and fingerprints. The advantage is that it is convenient for collection.In this paper, the finger veins are segmented, binarized, and translated, and the dimensionality is reduced by compressing the feature vectors. At the same time, a good recognition rate is achieved.
Keywords: Finger vein recognition SIFT algorithm pattern recognition image segmentation
基金:
引用
No.****
同行评议
勘误表
SIFT特征提取的指静脉认证技术研究
评论
全部评论0/1000