Please use this identifier to cite or link to this item: http://hdl.handle.net/10266/1763
Title: Generalized Solution to Fetch Information from the Cheques of Different Banks
Authors: Jain, Nikita
Supervisor: Kumar, Rajiv
Keywords: OCR;MICR code;resizing
Issue Date: 18-Jul-2012
Abstract: As the population is growing, bank transactions which involves the use of bank cheques is increasing rapidly. A great deal of work has been done on automatic processing of bank cheques like automatic extraction of items from cheque, signature verification, recognition of date from the cheque, amount filled by the user etc. Till now, the MICR code processing is performed using special type of scanner which reads the magnetic ink particles. These machines are very precisely designed which make them very expensive and requires constant maintenance to maintain their reliability. Moreover the MICR code is written with the special type of magnetic ink which is very costly. Any damage to MICR code characters will hamper its performance. It has been analyzed that no other approach has been developed for the processing of MICR code, so an effort has been made by the authors to develop a novel approach for processing the MICR code. The concept of OCR process serves as the basis for this approach. This approach automatically segments and recognizes the MICR code present at the bottom of the cheque. After analyzing number of cheques, it was observed that dimensions of digits were varying across the cheques. This made it difficult to find the pattern which helps to recognize the digit. In order to overcome this problem an approach was followed to manipulate the size of each digit to a specific dimension using the Resizing concept. MICR band which is present at the bottom of the cheque contains information. One of the information is MICR code. MICR code gives the important information about the name of the city, bank name and branch name of the bank who has issued that particular cheque. The uniqueness of our approach lies in the fact that it doesn’t necessitate any prior information and requires minimum human intervention. The system performance is quite promising on a large dataset of real cheque images.
Description: M.Tech. (Computer Science and Applications)
URI: http://hdl.handle.net/10266/1763
Appears in Collections:Masters Theses@CSED

Files in This Item:
File Description SizeFormat 
1763.pdf1.28 MBAdobe PDFThumbnail
View/Open


Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.