Choroidal Neovascularization (CNV) is a severe retinal disease characterized by abnormal growth of blood vessels in the choroidal layer. Current diagnosis of CNV depends mainly on qualitative assessment of a temporal sequence of fundus fluorescein angiography images. Automated segmentation and identification of the CNV lesion types (either occult or classic) is required to reduce the inter-and intra- observer variability and also to reduce the manual segmentation effort and time. In this work, we present automatic segmentation method for the CNV lesions. The method is based on developing a novel model to describe the temporal intensity variation of the image sequence. The model parameters at each pixel are used to construct a feature vector that is used to classify the different pixels into areas of classic CNV, occult CNV and background. Preliminary results on four datasets show the potential and effectiveness of the method to segment and identify the different types of CNV lesions.