Number Plate Recognition using Analytical Fourier Mellin Transform

  • Anshul Mittal
  • Mayank Sharma

Abstract

An automated, robust and fast vehicle license plate recognition system has become a essential necessity for traffic control and law enforcement in the present scenario. With increasing volumes of traffic being added to Indian roads every year and rough weather conditions prevalent, the challenge has become tougher. Here, we present a novel OCR-based method for number plate recognition using the Analytical Fourier Mellin transform which is fast and yields better results as compared to basic Fourier Mellin Transform.

The overall system uses an intelligent but cost-effective camera, which also uses minimal space, because the processor and the sensor are integrated in one package. They are more robust and reliable in outdoor environments, they require less maintenance and they are well suited for cost sensitive applications. Traditional correlation methods have to calculate the correlation over all possible rotations to determine the rotation angle between the template and query number plate character image when this angle is unknown. In this paper we achieve the alignment by Analytical Fourier-Mellin transform (AFMT), which can quickly align two images without re sampling errors and the high computation cost involved in conversion from log-polar to Cartesian. The Signature Technique has been used for the Number Plate Localization and Peaks to Valleys method have been used for character segmentation. As in the Indian Scenarios, the character distortion caused by dirt particles, image blurring or shadows that partially cover the license plate may cause problems. Thus, the segmentation is preferably parameterized to be sensitive to spaces. To further speed up the system, the property of symmetry of the vehicle rear and front parts are uniquely used to make the process of localization/ identification of the license plate quicker.. To ensure the identified character is not a caption a syntax analysis is subsequently performed. The experimental results with number plate correlation databases show that the proposed algorithm is effective.

Keywords

Radar 

Preview

Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.

References

  1. [1]
    D. Casasent, D. Psaltis. “Scale invariant optical transform,” Optical Engineering, 15(3):258- 261, May/ June 1976.Optical Engineering, 15(3):258- 261, May/ June 1976.Google Scholar
  2. [2]
    T. Yatagay, K. Choji, H. Saito. “Pattern classification using optical Mellin transform and circular photodiode array, Optical Communication, 38(3):162- 165, August 1981.Google Scholar
  3. [3]
    Y. Sheng, H. H. Arsenault. “Experiments on pattern recognition using invariant Fourier-Mellin descriptors,” J. of the Optical Society of America A, 3(4):771- 776, June 1986.Google Scholar
  4. [4]
    A. E. Grace, M. Spann. “A comparison between Fourier- Mellin descriptors and moment based features for invariant object recognition using neural networks,” Pattern Recognition Letters, 12:635- 643, October 1991.Google Scholar
  5. [5]
    P. E. Zwicke, Z. Kiss. “A new implementation of the Mellin transform and its application to radar classification,” IEEE trans. on Pattern Analysis and Machine Intelligence, 5(2):191- 199, March 1983.Google Scholar
  6. [6]
    F. Ghorbel. “A complete invariant description for gray-level images by the harmonic analysis approach,” Pattern Recognition Letters, 15:1043- 1051, October 1994.Google Scholar
  7. [7]
    R. Milanese, M. Cherbuliez. “A rotation-translation-scale-invariant approach to content-based image retrieval,” J. of Visual Communication and Image Representation, 10:186- 196, June 1999.Google Scholar
  8. [8]
    R. Bracewell. The Fourier transform and its applications, McGraw-Hill Inc., 2nd edition, New York (USA), 1986.Google Scholar
  9. [9]
    F. Ghorbel. “Towards a unitary formulation for invariant image description: Application to image coding,” Annals o Telecommunication, 53(5/ 6): 242- 260, May/ June 1998.Google Scholar
  10. [10]
    P. E. Anuta. “Spatial registration of multispectral and multitemporal digital imagery using fast Fourier transform techniques,” IEEE Trans. on Geo. Elec., 8:353- 368, 1970.CrossRefGoogle Scholar
  11. [11]
    D. Casasent, D. Psaltis. “Position, rotation, and scale invariant optical correlation,” Applied Optics, 15:1795- 1799, 1976.Google Scholar
  12. [12]
    D. Casasent, D. Psaltis. “Space-bandwidth product and accuracy of the optical Mellin transform,” Applied Optics,16:1472, 1977.Google Scholar
  13. [13]
    D. Casasent, D. Psaltis. “Deformation invariant, spacevariant optical pattern recognition,” In Wolf E (Ed): Progress in Optics, North-Holland Publishing Co, Amsterdam, 290- 356, 1978.Google Scholar
  14. [14]
    A. S. Jensen, L. Lindvold, E. Rasmussen. “Transformation of image positions, rotations, and sizes into shift parameters,” Applied Optics, 26:1775- 1781, 1987.CrossRefGoogle Scholar
  15. [15]
    S. Derrode, F. Ghorbel. “Robust and efficient Fourier-Mellin transform approximations for graylevel image reconstruction and complete invariant description,” Computer Vision and Image Understanding, 83(1):57- 78, 2001.MATHCrossRefGoogle Scholar
  16. [16]
    www.tn.gov.in/sta/regmark.htmGoogle Scholar
  17. [17]
    www.morth.nic.inGoogle Scholar
  18. [18]
    Canny, J, A Computational Approach to Edge Detection , IEEE Transactions on Pattern Analysis and Machine Intelligence, vol. 8, pp. 679-698, 1986.Google Scholar
  19. [19]
    S. Goh. “The Mellin transformation: Theory and digital filter implementation,” Ph.D. dissertation, Purdue University, West Lafayette, I.N., 1985.Google Scholar

Copyright information

© Springer Science+Business Media B.V. 2008

Authors and Affiliations

  • Anshul Mittal
  • Mayank Sharma

There are no affiliations available

Personalised recommendations