This paper presents the algorithm to extract the text from the degraded and distorted images based on pattern recognition techniques used in image processing. Degraded images are quite often observed in old manuscripts, where it is strongly needed to enhance the quality of the manuscript for reproduction efficiently. The degradation of text may arise due to oxidation of ink used for printing/writing of the manuscript or extra ink spots or somewhere inkless printing. Degradation may also arise from the image grabbing techniques from a good quality image. This may arise due to poor illumination or low resolution of the camera being used. However, distortion in text may arise on account of incomplete character printing or deletion of some part of a character due to aging or any other reason. Therefore, to reproduce the text from the old manuscripts, it is required to apply some non-destructive techniques (NDT) to extract the useful text. Pattern recognition in image processing provides the solution for this task.

1. I.Guyon, A scaling law for the validation-set training-set size ratio, AT&T Bell Laboratories, 1997.
2. A.Souza and al., Automatic filter selection using image quality assessment, 7th ICDAR Conference, 2003.
3. Hybrid Image Thresholding Method using Edge Detection, Febriliyan Samopa† and Akira Asano, IJCSNS International Journal of Computer Science and Network Security, VOL.9 No.4, April 2009.
4. Digital Image Processing - Gonzalez,Woods, Edition 1997
5. A Versatile Machine Vision System for Complex Industrial Parts (IEEE Trans. On Computers, Sep. 1977, No. 9, Vol. C-27).
6. " Fuzzy Sets, Uncertainty and Information " A book by George J. Klir & Tina A. Folger.
7. NI Vision Based Automatic Optical Inspection (AOI), Proceedings of 2009 IEEE International Conference on Applied Superconductivity and Electromagnetic Devices Chengdu, China, September 25-27, 2009.
8. "Neural network And Fuzzy Systems – A Dynamical System Approach To Machine Vision Intelligence" Edition-1990, A book by Bart Kosko.
9. "Fuzzy Logic And Control: Software And Hardware Applications" A book by Mohammad Jamshidi, Nader Vadice and Timothy j. Ross.
10. "Edge Moments in Pattern Recognition" Thesis by Dr. H.K. Sardana.

Ndt Cog