Image Segmentation Using Two Step Splitting Function

Image processing and computer vision is widely using Level Set Method (LSM). In conventional level set formulation, irregularities are developed during evolution of level set function, which cause numerical errors and eventually destroy the stability of the evolution. Therefore a numerical remedy called re-initialization is typically applied periodically to replace the degraded level set function. However re ?initialization raises serious problem that is when and how it should be performed and also affects numerical accuracy in an undesirable way. To overcome this drawback of re-initialization process, a new variation level set formulation called Distance regularization level set evolution (DRLSE) is introduced in which the regularity of the level set function is internally maintained during the level set evolution. DRLSE allows more general and effective initialization of the level set function. But DRLSE uses relatively large number of steps to ensure efficient numerical accuracy. Here in this thesis we are implementing faster and equally efficient computation technique called two step splitting method (TSSM). TSSM is physio-chemical reaction diffusion equation in which firstly LSE equation get iterated and then regularize the level set function from the first step to ensure the stability and hence re-initialization is completely eliminated from LSE which also satisfy DRLSE.