Vehicle Detection System Using SVM Classification and HAAR Filter

Vehicle detection is very much important in avoiding accidents and for traffic monitoring. Various features such as colors, edges are used for vehicle detection. Gaussian mixture models (GMM) is used for background removal. The method of detection consists of training phase and detection phase. In both the training phase and detection phase we are using the same features for extraction. Afterwards the extracted features are used to classify whether it is a vehicle pixel or a non vehicle pixel using SVM. To improve the detection we are using the haar features. Detection in this approach is solenly based on pixels.