Face image relighting method

Abstract

The invention discloses a face image relighting method, which comprises the steps of: performing face recognition on a two-dimensional face image, and acquiring two-dimensional feature points; performing face three-dimensional reconstruction according to the two-dimensional feature points and three-dimensional face model data, and generating a three-dimensional grid of a three-dimensional face image; carrying out local illumination rendering on the three-dimensional grid according to the illumination rendering parameters and a pre-established illumination model, and generating a first two-dimensional face illumination mapping graph; and performing skin color detection on the two-dimensional face image, and acquiring a two-dimensional face image skin color probability value; performing non-skin-color region protection on the first two-dimensional face illumination mapping graph according to the skin color probability value, so as to obtain a second two-dimensional face illumination mapping graph; carrying out edge extension on a facial contour of the second two-dimensional face illumination mapping graph, so as to obtain a third two-dimensional face illumination mapping graph; and subjecting the third two-dimensional face illumination mapping graph and the two-dimensional face image to relighting fusion, and generating a lighting result map. The face image relighting method provided by the invention is simple, feasible, fast and efficient, and is high in user controllability.

Description

technical field [0001] The invention relates to the technical field of digital image processing, in particular to a method for re-illuminating a human face image. Background technique [0002] In recent years, portrait beauty and makeup applications have become popular among young users. In order to obtain good shooting results, users usually need to be in a good lighting environment and lighting conditions. Due to the lighting conditions and the position of the light source, it is easy to lose the three-dimensional details of the face, resulting in a flat face after makeup, which lacks a sense of reality. As an improvement to the effect of the above-mentioned applications, a method of performing facial relighting (Relighting) on ​​the portrait photos in the later stage is proposed, that is, a method of relighting the face image. The method is to adjust the face image of the target person according to the change of illumination or external environment, and generate a face i...

AI Technical Summary

Problems solved by technology

[0004] From the above methods, it can be seen that the existing face image re-illumination technology needs to prepare a large amount of sample data in different face poses in advance.

At the same time, considering that different races and skin colors have different facial structures (face shapes) and different light reflectances, this greatly increases the number of samples required for preparation

Based on the interference of these environmental factors and the technical solutions adopted, it is difficult for the above-mentioned methods to accurately reconstruct and restore the 3D structure of the face

At the same time, due to the skin color and different facial structures, it is also difficult to find out the exact position of the light source and the correlation coefficient of the material

In short, the existing technical solutions are more complicated, difficult to master, and less effective in implementation

Images