Automatic lip contour extraction using both pixel-based and parametric models

Lip segmentation, a sub-step of facial analysis, becomes more and more important in the fields of pattern recognition and human-computer interaction. In this paper, a novel algorithm for lip contour extraction that combines the merits of the pixel-based model and the parametric model is presented. Lip corner detection is the first step of this algorithm, utilizing a three-curve lip model to describe lip contour. The improved "Jumping Snake" algorithm is used to extract feature points for lip model implementation, which makes the novel algorithm more accurate and flexible. It is also made more robust with the introduction of geometric constraints. Experiments show that this approach with geometric constraints provides satisfactory results. It can also serve as an initial attempt for further applications.