Fast and Improved Level Set Method for Image Segmentation

In recent years, image segmentation algorithms are widely used for segmenting the image into different parts which can be used for further analytical purposes. Nowadays, many applications such as iris segmentation for authentication, brain tumor detection and number plate segmentation uses different segmentation methods. Among them level set methods emerged widely due to its efficiency in handling the change in the topology of images. For solving the level set equation, the conventional level set methods uses some finite approximations schemes for providing numerical stability. These schemes take lot of time for the curve to evolve. This problem is tackled using the lattice Boltzmann method (LBM). In this method, a fuzzy energy function based on fuzzy c-means objective function is minimized using the gradient descent method to create the level set equation and LBM is used to solve the level set equation in order to provide parallel programming. The proposed method is insensitive to the initial position of the contour. This method easily detects brain tumor, segmented iris and character from number plate of vehicle is selected.