Active Segmentation Model Based On The Shape Of A Priori Information

In the case of noise, occlusion or strongly clutter background, it's always difficult to segment the object correctly. One effective way to solve this problem is to utilize the priors. In this paper, we propose an active contour algorithm incorporating with prior shapes. We design an energy functional composed by two complementary terms. The first one contains the image information, including image gradients and the homogeneous of region intensity. The second term is based on the shape information, which constrains the active contour to approach a shape compatible with a statistical shape model of the shape of interest. In order to capture the shape information, level set method is used to represent the shape. Then the probability distribution is defined and principle component analysis is applied to capture the variance mode of shape. Finally we combine the shape information with active contour model. In this way, the segmentation based on the information is not only from the image, but also from the training set of shape. The segmentation of the object of interest is given by the minimum of our energy functional. This minimum is computed with the calculus of variations and the gradient descent method, which provide a system of evolution equations solved with the well-known level set method. Results are demonstrated on synthetic data and medical imagery.