In this paper a Vehicle Identification and Authentication System is developed for traffic monitoring. To prevent unauthorized vehicles from entering the private areas, vehicle based authentication technologies are employed. The captured color image of the vehicle is converted to gray scale image. Gray scale image is converted into binary image using Sliding Concentric Windows (SCW) method. Pixels are labelled into components based on pixel connectivity using Connected Component Analysis (CCA) technique. The labelled components are examined and detected license plate is processed to isolate characters. These characters are sent to Probabilistic Neural Network (PNN). PNN uses a supervised training set to develop distributed functions within a pattern layer. Vehicle entering an area is considered as authenticated if it is registered. Registered vehicles’ number plate indicates the state and district to which particular vehicle belongs to. Vehicles are considered to be authenticated, if it belongs to a particular state and district, after which the vehicles’ type and other details are extracted from the database.