Research On 3D Face Expression Synthesis Based On Depth And Color Image

Facial expression is the most intuitive expression of human’s inner mood.Therefore,it can achieve the purpose of expressing people’s different feelings through synthesize different facial expressions artificially.While the 3D facial expression with high authenticity can simulate the emotion expression of human more meticulously.3D face expression synthesis already occupy a very important position in the film special effects production,human-computer interaction,game production,video conferencing and other applied research areas.In this paper,3D face expression synthesis method is divided into three parts.First,we use Kinect to collect color and depth image of human face.The Kinect color camera and depth camera coordinates are calibrated prior to acquisition.In the process of acquisition,the face is identified and tracked,and the coordinates of feature points are extracted.Then FaceGen 3D face model is selected as the general face.According to the acquired facial feature point coordinates,the coordinates of the feature point and the non-feature point in the grid model are adjusted to the shape of the specific face.The synthetic full-angle face texture image is vertically mapped to the adjusted mesh model using OpenGL combined with the texture linear interpolation method to generate a specific 3D face.Finally,based on the facial action AU defined in the expression coding system FACS,the vertices in the model are adjusted.Re-mapping the 3D face model to generate a 3D face with a specific expression.In this paper,the face of the author as the prototype of 3D face modeling.Six different facial expressions can be synthesized by user selection,including joy,sadness,surprise,anger,fear and disgust.The features of the six basic three-dimensional facial expression details are highlighted in this paper.Which has a high similarity comparison with the real facial expression.And the time efficiency is significantly improved compared with the facial expression synthesis method based on the human face three-dimensional scanning data and motion capture technology.Comprehensive analysis and comparison of this method and other methods Experimental results are summarized in this paper the advantages of synthetic 3D facial expression design methods.