Gait identification using shadow biometrics

摘要: We propose a novel biometrics method based on shadows (shadow biometrics, SB) and introduce SB-based person identification for vision-based surveillance system. Conventional biometric body movements, as is the case in gait recognition, uses cameras that provide good view of entire human body. Aerial search systems only see from top with smaller cross-section less details than seen side views, which further aggravated by lower resolution associated this imagery. Shadows, i.e. projections due to Sun, or artificial lights at night, can offer information cannot be directly view. In paper we use SB identification, automatically extracting captured video images, processing them extract features, analyzed spherical harmonics. demonstrate shadow-based experiments inside building using light outside under Sun. The introduced harmonics outperforms methods Fourier transform, energy image, active image. Furthermore, show combination shadow areas, an oblique camera upper floor building, has better performance information.