Automatic Vehicle Plate Recognition (AVPR) is the extraction of vehicle license plate information from an image or sequence of images. From the past thirty years, AVPR is becoming the challenging and interesting area of research. AVPR systems include a wide range of applications. Numerous real-world applications such as electronic toll collection, automatic parking management, access control, radar-based speed-control, border control, criminal pursuit, traffic law enforcement, etc. have been benefited from it. A lot of commercial AVPR systems are available today and yet there are many challenges and issues in accurate recognition of license plates. In India, number plate standards are rarely practised. Licence plates recognition has many problems like unnecessary text, different font size and font type, blur, skew, environmental factors etc. The variations of the licence plate types or environments cause challenges in the recognition of number licence plates. The major objective of this thesis is to develop a robust, accurate and reliable automatic vehicle license plate recognition system. Our suggested approach is performed in three phases: In the first phase, the input image is pre-processed. Character regions are extracted in the second phase, and in the third phase, recognition of extracted characters is performed. The present work has been performed to recognize Indian licence plates