International Journal of Science, Engineering and TechnologyAuthors: Assistant Professor G. Sudheer kumar, K. Bhagyasree
Abstract: Bank cheque verification is a critical banking operation that requires high accuracy to minimize fraud and processing delays. This study introduces an intelligent cheque verification framework that combines image processing with deep learning techniques to automate the authentication process. The proposed approach extracts essential cheque details such as the IFSC code, cheque number, account number, handwritten amount, printed text, and account holder's signature from scanned cheque images. Image preprocessing techniques are applied to enhance image quality and isolate important regions before analysis. Optical Character Recognition (OCR) is employed to recognize printed characters, while a Convolutional Neural Network (CNN) is utilized for identifying handwritten numerical values. Signature verification is performed using Scale-Invariant Feature Transform (SIFT) for feature extraction and a Support Vector Machine (SVM) for classification. The integrated framework improves verification reliability, accelerates cheque processing, and reduces manual intervention. Experimental evaluation on a standard cheque image dataset demonstrates that the proposed system achieves high recognition accuracy for text, handwritten digits, and signatures, making it a practical solution for secure and efficient banking applications.
