Automatic Car license plate Recognition system using Multiclass SVM and OCR

  IJETT-book-cover  International Journal of Engineering Trends and Technology (IJETT)          
  
© 2015 by IJETT Journal
Volume-30 Number-7
Year of Publication : 2015
Authors : Ravindra Madhukar, Ravendra Ratan Singh
  10.14445/22315381/IJETT-V30P269

MLA 

Ravindra Madhukar, Ravendra Ratan Singh"Automatic Car license plate Recognition system using Multiclass SVM and OCR", International Journal of Engineering Trends and Technology (IJETT), V30(7),369-373 December 2015. ISSN:2231-5381. www.ijettjournal.org. published by seventh sense research group

Abstract
Automatic car license plate recognition system has always attracted researchers. It is a dynamic region of exploration in machine vision and its application. Over the years there have been many techniques where in car license plate recognition systems have been successfully proposed and developed. Broadly the car license plate recognition systems are classified as template matching based and extracting features based. Template matching based is simple and straight forward method but it is vulnerable to any font change, rotation and noise. Extracting feature based method is a fast method and more accurate but feature extraction is a challenge and any no robust feature decreases the recognition accuracy. On the basis of my preliminary results I propose an integrated template and feature based method for automatic car license plate recognition system for INDIAN cars license system. I aim in developing an automatic car license recognition system based on still images. Image database set is collected for different categories of car license system adopted in INDIA. Template matching is done via implementation of optical character recognition system which shall help in recognizing characters of the license plate. But to enhance the speed and to increase the accuracy of the system the images are classified using a new variant of state vector machine known as Multiclass SVM. The idea is to implement the proposed system using the computational intelligence concept, image processing concept and artificial intelligence concept. The proposed system is then evaluated via MATLAB’s Computer Vision Toolbox and Artificial Intelligence toolbox.

 References

[1] Moshe Blank, Lena Gorelick, Eli Schechtman, Michal Irani, and Ronen Basri.Actions as space time-shapes. In IEEE International Conference on Computer Vi-sion, ICCV, 2, pp.1395–1402, Oct 2005.
[2] J. Yamato, J. Ohya, K. Ishii, Recognizing human action in time-sequential images using hidden Markov model, Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition. 379 – 385, 1992.
[3] S. Haykin, Neural Networks: A Comprehensive Foundation, Prentice Hall, 1998.
[4] Standard MPEG1: ISO/IEC 11172, Coding of moving pictures and associated audio for digital storage media at up to about 1.5 Mbit/s, 1996.
[5] T. Kohonen, The self-organizing map, Proceedings of the IEEE 78 (9):1464 – 1480. 1990.

Keywords
computational intelligence, image processing and artificial intelligence.