Rotation immunity characteristics of non-contact fingerprint recognition using local luminance variation

This study presents the non-contact type of fingerprint recognition method using the local features: the histograms of oriented gradients (HOG) or local luminance variation. A non-contact type of fingerprint recognition can solve contact type problems that the performance decreases by surface condition and so on. However, the noncontact fingerprint capturing causes low contrast between the ridges and the valleys and induces the 3-D rotation of the fingerprint image. In this study, the fingerprint recognition performances of the proposed method were investigated using the fingerprint images rotated in the yaw rotation. From the experimental results, the local luminance variation had the immunity of yaw rotation variation from -10° to 10°.