Fast Facial Beautification Algorithm And System Based On Edge-preserving Smoothing Filter And Edit Propagation

Old saying goes- The heart of beauty, everyone. Either paintings of ancient writers or digital photos show that human never stop their footsteps pursuing beauty. Maybe its concept is too abstract, the essence of beauty always does not have a quantifiable criteria. However, thanks to modern computer processing power, as well as advances in digital image processing technologies, it was discovered that some factors which affect the beauty and was successfully applied to improve the beauty of face portrait.On the basis of previous research, this thesis proposes a fast algorithm for face beautification. Considering effectiveness and efficiency of the beautification processing, this thesis improve the performance of the algorithm so that it meets practical requirements. The main contributions of this thesis include:This thesis implement several edge-preserving smoothing filter while comparing their performance, and point out that guide filter is suitable for the layer decomposition algorithms, since it can preserve the edge of face in different layers while its productivity has risen by 20 times to others.This thesis study Image editing technology, and applied an Adaptive mask generation algorithm based on Edit-Propagation model to the layers enhanced. By comparing a variety of linear system solving algorithm, we proposed approximation multigrid algorithms which solve the higher dimensional linear systems with the efficiency rise by 4 times, making the mask generation meet real-time requirements.This thesis Implemented a piecewise affine face morphing, which get satisfactory results and improvement of efficiency nearly 20 times compare to commonly used MFFD algorithm when processing normal or high resolution images.Finally, this thesis implements a quick beautification system with outstanding effectiveness. Meanwhile, to facilitate technology diffusion, this complete set of algorithms are packaged as a portable algorithms library that can be used to other applications. In addition, based on algorithm interface provided by the library, this article implements several beautification applications on the PC under WPF. In collaboration with the author’s lab, the beautification system was deployed in Microsoft’s Windows Azure cloud services platform, to provide face beautification services for the outside world.As mentioned above, the algorithm for face beautification in the academic has considerable research value, at the same time, it has broad application prospects in real life, non-professional users can use facial beautification system image beautification, and the system can also provide aid for graphic designers while designing Virtual characters.