Real-Time Facial Sketch Synthesis from Video

In this paper, a facial sketch system based on real-time video is presented. This system automatically synthesizes facial edge sketch by tracking facial main control points (MCPs) accurately. There are four key steps in the system. First, dozens of main control points are obtained from frontal face image by Active Appearance Model (AAM). Second, when face moved or rotated, the new positions of MCPs are computed with optical flow algorithm. Third, face rotate matrix is obtained by considering spatial coherence and time coherence of MCPs. In the last step, cubic B-spline interpolation is introduced to connect the MCPs smoothly for sketch synthesis. Experiments shows our system has good real-time performance, and can get good results when face moved or rotated.

[1]  J.-Y. Bouguet,et al.  Pyramidal implementation of the lucas kanade feature tracker , 1999 .

[2]  Timothy F. Cootes,et al.  Constrained active appearance models , 2001, Proceedings Eighth IEEE International Conference on Computer Vision. ICCV 2001.

[3]  Hong Chen,et al.  A high resolution grammatical model for face representation and sketching , 2005, 2005 IEEE Computer Society Conference on Computer Vision and Pattern Recognition (CVPR'05).

[4]  Dimitris N. Metaxas,et al.  Emblem Detections by Tracking Facial Features , 2006, 2006 Conference on Computer Vision and Pattern Recognition Workshop (CVPRW'06).

[5]  Nanning Zheng,et al.  Example-based facial sketch generation with non-parametric sampling , 2001, Proceedings Eighth IEEE International Conference on Computer Vision. ICCV 2001.

[6]  Takeo Kanade,et al.  An Iterative Image Registration Technique with an Application to Stereo Vision , 1981, IJCAI.

[7]  Jiebo Luo,et al.  Accurate Dynamic Sketching of Faces from Video , 2007, 2007 IEEE Conference on Computer Vision and Pattern Recognition.

[8]  Timothy F. Cootes,et al.  Active Shape Models-Their Training and Application , 1995, Comput. Vis. Image Underst..