Detect Fraud Bank Checks with Convolutional Neural Network Processing Algorithm

Abstract

Abstract

The purpose of this study is to design and build a fraud check detection system with a Convolutional Neural Network (CNN) algorithm to quickly and easily detect a fraud or altered bank check in real time. A MATLAB Deep Learning Toolbox with related CNN algorithm is used to assist this study to quickly detect any fraud or altered bank check when it is deposited and scanned by a scanner in any ATM in US. The testing and validation processes have been performed to confirm the effectiveness and correctness of this detecting system. The current correct detecting rate is 97.5% for checks deposited via ATMs.

Keywords: Altered check inspections, convolutional neural network processing, fraud bank check detections, image signal processing, image detections