Optical Character Recognition in Banking Sectors Using Convolutional Neural Network

Optical Character Recognition is to changing computerized pictures or content into machine editable content and also recognizes the text within a digital image. It is very useful in banking sectors. Till now, Bank deposit slips are processed manually every day in all the banks, it takes a long time. The proposed approach is to recognize the handwritten digits and characters from the bank deposit slip using deep learning techniques. It makes all the banking processes and transactions faster. The handwritten digits and characters such as account holder name, account number and date are segmented by image processing techniques. The segmented character and digits are fed to the CNN model. The model recognizes the digit and character from the deposit slip. The Proposed model delivers better recognition accuracy.