Abstract: Minimum level of the user expertise required for professional software to achieve the enviable results. In this project, an algorithm is used to identify facial flaws / imperfection. Any such algorithm would be agreeable to facial retouching applications. The identification of flaws / imperfections can allow these skin textures to be processed in a different way than the surrounding skin without much user contact. For identification, Gabor filter responses along with texture orientation field are used as image features. A bimodal Gaussian mixture model (GMM) represents distributions of Gabor features of normal skin versus skin imperfections. Then, a Markov random field model is used to include the spatial relationships among neighboring pixels for their GMM distributions and texture orientations. An expectation-maximization algorithm then classifies skin versus skin flaws/imperfections.Once detected, flaws / imperfections are removed completely instead of being blended or blurred. An exemplar-based constrained texture synthesis algorithm is used to in paint irregularly shaped gaps left by the removal of detected flaws / imperfections. Results are conducted on images downloaded from the Internet or taken from the system to show the efficacy of algorithms.

Keywords: Exempler based synthesis algorithm, Matlab, GMM.