Extraction of Bank Cheque Fields Based on Faster R-CNN
Hakim A. Abdo, AHMED A. ABDU, Ramesh R. Manza, Shobha Bawiskar · Advances in intelligent systems research/Advances in Intelligent Systems Research · 2023
The cheque field extraction is a critical step in automating bank cheque processing and is the first step in implementing a cheque recognition system.Many approaches for extracting the bank cheques components have been suggested.However, the complexity of the backdrop, the design variety of bank cheques, the variety of font sizes, and different patterns of writing remain a difficulty that necessitates the employment of precise algorithms.In this paper, we present a novel approach to extract the bank cheque components, in presented approach we used an innovative model called Faster R-CNN.This model represents the pinnacle of object recognition since it eliminates the need to manually extract image features and instead segments images to provide candidate region suggestions automatically.The IDRBT Cheque Image Dataset is used to train and test the Faster R-CNN model.The findings demonstrate that the model is capable of properly detecting the bank cheque fields.The extraction of bank cheque fields using Faster R-CNN achieves an accuracy of 97.4%, which outperforms other techniques.