As the variety of algorithms for person recognition continues to enrich, 3D face recognition is expected to expand

At present, for the main challenges in the field of face recognition light perception, enriching algorithm capability has become the competitive focus of algorithm vendors.

Due to the size of cell phones, there is limited space for upgrading the optical system as well as the sensing system in the optical perception system. Therefore, in the field of face recognition, algorithms become the main way to make up for the hardware limitations of the optical perception system and optimize the face recognition function. For example, algorithms such as histogram equalization, camera distortion correction, median filtering, etc., enable face recognition capability to be upgraded continuously.

In addition, 3D face recognition is expected to develop and advance in the face recognition market in the future. Since 2D face recognition has been researched for a longer time and there are more related algorithms, the current face recognition solutions on the market are mainly 2D solutions. However, the 2D information lacks depth information and cannot completely express the real face, which leads to the limited accuracy rate of 2D face recognition. Compared with 2D face recognition, 3D face recognition has two more links of alignment and 3D reconstruction, and can also obtain depth information through structured light, TOF or binocular stereo vision, which can reflect the face more realistically and thus can effectively improve the accuracy rate of recognition. With the progress of manufacturers' R&D technology, the penetration rate of 3D solutions in cell phones will further increase.