IJSRSET1623213 | Received: 20 June 2016 | Accepted : 24 June 2016 | May-June 2016 [(2)3: 793-800] © 2015 IJSRSET | Volume 1 | Issue 6 | Print ISSN : 2395-1990 | Online ISSN : 2394-4099 Themed Section: Engineering and Technology 793 Real Time License Plate Recognition via Watershed and Viterbi Algorithm Kaushalya Bakde , Suyash Agrawal Computer Science & Engineering Department, Rungta College of Engineering & Technology, Kurud Kohka Bhilai, Chhattisgarh, India ABSTRACT License Plate Recognition (LPR) is the extraction of vehicle license plate information from still images or frame sequences (videos). Character segmentation & recognition has long been a critical area of the OCR process. The characters are detected in the order defined by the matching quality. In this paper three main procedures watershed, thresholding and hidden markov model based Viterbi algorithm was used to perform license plate segmentation and recognition tasks. The watershed transformation with thresholding algorithm based on the gradient approach gives good results for segmentation of characters. This is mainly designed for Indian Car license plate. The procedure follows a simple and effective way to segment and recognize the characters. This paper also presents extensive experiments using real video sequences to verify the proposed method. Keywords: License plate recognition (LPR), Watershed algorithm, Thresholding, Hidden markov Viterbi algorithm, Optical Character Recognition (OCR) for cars. I. INTRODUCTION Segmentation and recognition are two important tasks in image processing and computer vision. License Plate Recognition (LPR) is an integral part of Intelligent Transportation Systems (ITS). License plates are used for identification of vehicles all over the nations so it is illegal for two vehicles to have the same license number. License plate recognition (LPR) algorithms in images or videos are generally composed of the following three modules: (1) license plate detection, (2) character segmentation, and (3) recognition. The first two steps incorporate image processing techniques on still images or frame sequences (videos), whose evaluation relies on the true recognition rate and the error recognition rate. Among these steps, a very critical step is the license plate segmentation, which directly affects the overall system performance. A broad range of methods for license plate segmentation have been reported in the literature. Vehicle registration plates are formatted as follows:  Plates for private car and motorized two-wheeler owners have black lettering on a white background (e.g., CG.18.K.1316).  Commercial vehicles such as taxis and trucks have a yellow background and black text  (e.g., MH.12.CD.2544).  Commercial vehicles available on rent for self-drive have yellow lettering on a black background  (e.g., KA.03.AB.8192).  The President of India and state governors travel in official cars without license plates. Instead they have the Emblem of India in gold embossed on a red plate. Some Indian number plates are shown below: Figure 1. Samples of license plates in India