Key Image Processing Techniques Applied In Virtual Facial Plastic Surgery

Along with the development of our society, people's standpoint towards plastic surgery has changed. More and more people want to obtain ideal appearances by doing plastic surgeries. Facial plastic surgery is a clinic cure technique with high venture which requires high accuracy. The failure of a facial plastic surgery will do great harm to patient both physically and psychologically. This problem is wildly concerned by the world. Conventional facial plastic surgery cannot evaluate the facial morphological parameters objectively and quantificationally before the surgery. And it cannot show the surgery effects beforehand either. Thus, the communication between the surgeons and the patients is not well enough to make the cure projects accord with the requirements of patients which might cause a failure plastic surgery. Contrasted with the conventional facial plastic surgery, the Virtual Facial Plastic Surgery (VFPS) system which is created according to virtual reality surgery technique can evaluate the surgery venture and visualize the surgery effect preoperatively.This paper details some key techniques in digital image processing of the VFPS system, such as facial features location, eye detection, image mosaic, image fusion, image morphing and preoperative visualization etc. The facial features coming from digital cameras were located. Sixteen feature points of the frontal human face photos and six feature points of the lateral face photos were located. A hybrid projection function has been improved to do eye locations in this paper. The coefficients in the function were discussed by experimental method. A test of 50 images in BioID face database has been done with the accuracy of 96% and the average relative error of 0.0873. In PSFace database with 20 images, the accuracy was 100% and the average relative error was 0.0855. A trapeziform edge information based algorithm for image fusion was used to do frontal nose image mosaic in this paper. The image morphing technique was used to achieve preoperative visualization for lateral nose images. By experiment, the techniques were proved to be efficient. The result will help to improve the VFPS system and the facial plastic surgery.