Effects of meteorological factors on finger vein recognition

Abstract

Finger vein recognition is a biometric method utilizing the vein patterns inside one's fingers for personal identification. Current recognition algorithms use completely image-based feature extraction and matching technology. However, we notice that the performance of the recognition system varies when users are exposed to different meteorological environments. Researchers have proven that many environmental factors may affect the vessel and blood flow, which could directly change our finger vein patterns. In this paper, we study the impact of different meteorological factors such as temperature, humidity, atmospheric pressure, and wind, on the state-of-the-art recognition algorithms. Based on the experiment results, we find that temperature is the most significant factor. To further study the influence of temperature, we first collect finger vein samples from a group of subjects under different environment temperatures, while keeping all the other factors fixed. Then we conduct experiments to provide evidence to confirm the impact of temperature difference on both the recognition algorithm and our recognition system. Finally, we propose two methods, dynamic template selection and threshold adjustment, to reduce such impact from temperature. The experimental result proves the effectiveness of our methods.