A long-distance 3D face recognition architecture utilizing MEMS-based region-scanning LiDAR

Abstract

In this paper, a novel 3D face recognition system utilizing the MEMS-based indirect Time-of-Flight (ToF) region-scanning LiDAR is proposed for long-distance person identification. Specifically, this face recognition system consists of two parts: (1) detection of the targeted face region by IR amplitude image and (2) 3D face recognition with the high-resolution face data of region-scanning. The proposed system is carried out on the self-collected dataset and gets maximum Rank-1 recognition rate of 95% in various distance and illumination conditions. Moreover, the proposed system outperformed the other 3D face recognition system with conventional ToF sensors in the aspect of the Rank-1 recognition rates at long distance of more than 3 meters.