International Conference on Information and Knowledge Technology

فارسی

Home / شانزدهمین کنفرانس بین المللی فناوری اطلاعات و دانش

A Real-Time and Robust Approach for Banknote Recognition

Authors :

Hani Abdi¹ Mohammad Javad Parseh²

1- دانشگاه دولتی جهرم 2- دانشگاه دولتی جهرم

Keywords :

Image Processing،ORB،SIFT،AKAZE،BRISK،Iranian Banknote

Abstract :

Automatic banknote recognition plays a critical role in real-world financial systems, especially for applications requiring real-time performance, such as assistive technologies for the visually impaired. This paper proposes a robust and lightweight pipeline that enhances keypoint-based feature extraction algorithms (ORB, SIFT, AKAZE, BRISK) through a tailored sequence of preprocessing techniques. The proposed approach is evaluated on a comprehensive dataset of 1117 images of Iranian banknotes across 29 classes, including worn and shadow-affected samples. Experimental results show a notable improvement in classification accuracy, G-mean, and runtime efficiency after preprocessing. The method achieves strong performance even on resource-limited systems, making it practical for embedded and real-time applications.