A novel approach for automatic extracting the user-entered data from bankchecks
Alessandro L. Koerich, L. L. Lee · 1999
Millions of handwritten or machine printed bankchecks have to be processed everyday. Since the bankcheck processing is merely a repetitional task it is desirable to realize it in an automatic fashion. Currently, thresholding 2;3;4 and image subtraction 1;6;7 techniques are being used for extracting the user{ entered data. The techniques based on image subtraction have shown more robustness to segment the user{entered data. Okada{Shridhar have suggested an approach based on a subtraction operation between one lled{in bankcheck image and the same bankcheck image without the lled{in information. This approach might not be feasible for real{life applications since a large database is needed to keep the sample images of un lled bankchecks. The main idea of this new approach is to handle only sub{images which contain the user{entered data. A template is used to extract these areas. For each of the ve resulting sub{images, the background pattern is eliminated by a subtraction operation, the baselines are detected and eliminated by using an algorithm based on the projection pro les and the character strings are eliminated by a subtraction operation between the sub{image and a generated binary image which contains every make{up character string. Finally, some pixels lost during the foregoing processing are recovered.