Image based classification of normal and special people using constrained local model

Detection and diagnosis of mental retardation in early stage is a problem for almost all Pediatric Doctors and Parents. The image database has been constructed by capturing face images of local normal and special persons and this work will help in early detection and diagnosis of Cognitive diseases. This paper focuses on implementation of Constrained Local Model (CLM) to classify the normal and special persons. CLM model uses two feature model, shape model and patch model to extract the desire features of facial images. CLM is based on shape model and patch model. The shape model is constructed using Principle Component Analysis (PCA). Patch model is constructed using Support vector machine (SVM). This paper is based on shape model. Features extracted from this approaches are used as input o supervised classifier such as Linear Discriminate Classifier (LDC) and Quadrature discriminate Classifier (QDC).