Face model reconstruction method and system, image processing system and storage medium

Abstract

The invention discloses a face model reconstruction method and system, an image processing system and a storage medium. The method and system are applied to a processing system with a plurality of near light sources and at least one image acquisition device. The method comprises the following steps: acquiring human face images acquired by an image acquisition device under the irradiation of each near light source, and selecting one or more human face images with least shadows as sample human face images; fitting a rough face model corresponding to the sample face image according to a face parameterization model database; dividing a skin area and a hair area in the rough face model; calibrating the position and the light source intensity of each near light source by utilizing a normal graphcorresponding to the skin area; segmenting a shadow area and a non-shadow area from the sample face image, and screening one or more pixels for realizing the non-shadow area from the calibrated nearlight sources; and performing photometric three-dimensional reconstruction according to the screened near light source to obtain an accurate face model. The system is simple in structure, low in costand simple and convenient to reconstruct, and calibration light sources do not need redundant auxiliary equipment.

Description

technical field [0001] The invention relates to the technical field of image processing, in particular to a face model reconstruction method, system, image processing system and storage medium. Background technique [0002] For 3D reconstruction of high-precision face models, the mainstream method is to use photometric stereo reconstruction. The principle is that the surface of the object has different intensities of reflection of light in different directions, and the magnitude of the reflection intensity is the angle between the light direction and the normal direction of the object surface. related. Therefore, by recording pictures of faces under different lighting conditions, a high-precision face model can be reconstructed. In general implementation, a high-power distant light source is used to generate approximately parallel directional light, which leads to a relatively large hardware system. In order to improve the portability of the system, it is necessary to use ...

AI Technical Summary

Problems solved by technology

[0003] In view of the shortcomings of the prior art described above, the purpose of the present invention is to provide a face model reconstruction method, system, image processing system and storage medium to solve the problem of high cost of high-precision three-dimensional face model reconstruction in the prior art

Images