Adaptive finger vein recognition framework with image quality analysis

• 作 者: LizhenZhou;GongpingYang;LuYang;YilongYin;XiaomingXi;
• 机 构: School of Computer Science and Technology,Shandong University,Jinan 250101,P.R.China
• 出 处: 2015全国理论计算机科学学术年会
• 发表日期: 2015年期

　　As one of the most promising biometric techniques, finger vein recognition has received increasing attention recently.In order to improve the recognition performance, a lot of efforts have been made on the feature extraction methods.However, these methods come with some challenges.For example,some methods are sensitive to noise and some methods are much time-consuming.Therefore, in order to overcome these challenges, we present an adaptive finger vein recognition framework with image quality analysis, which can effectively utilize these methods.In detail, when comparing two finger vein images,we firstly extract image quality features and calculate their quality difference.Then we use the quality difference as the auxiliary information to determine the feature extraction method.Finally, we compute the matching score to achieve the recognition result.Experimental results on the open HKPU database show that the proposed framework achieves desirable performance in both speed and recognition rate.