Performance Evaluation of Document Image Algorithms
- 353 Downloads
Performance evaluation for document image processing has a different emphasis than performance evaluation in other areas of image processing. Other areas of image processing can tolerate some error. Because it is so easily done nearly perfectly by humans, document image processing must also be done nearly perfectly. So the first aspect of performance evaluation for document image processing is to determine the domain in which the performance is nearly perfect. Outside this domain, the algorithm makes errors. Such instances of errors need to be examined and classified and categorized so that the weaknesses of the algorithm can be characterized.
KeywordsPerformance Evaluation Tuning Parameter Document Image Acceptance Test Surement Plan
Unable to display preview. Download preview PDF.
- 1.Abramovitz, M. and Stegum, I. (Eds.), Handbook of Mathematical Functions, Dover Publications Inc., New York, 1972.Google Scholar
- 2.Devijver, P. A. and Kittler, J., Pattern Recognition: A Statistical Approach, Prentice Hall International, London, l982.Google Scholar
- 3.Fununaga, K., Introduction to Statistical Pattern Recognition, Academic Press, New York, l972.Google Scholar
- 6.Johnson, N. and Kotz, S., Continuous Univariate Distribution-2, Houghton Miffin Co., Boston, 1970.Google Scholar
- 8.Pearson, E. S. and Hartley, E. O., Biometrical Tables for Statisticians, Vol. 1, (2nd Edition) Cambridge University Press, London, 1958.Google Scholar