Counterfeit detection of ringgit banknotes using image processing techniques

Citation

Salem Al-Hila, Turki Khaled (2023) Counterfeit detection of ringgit banknotes using image processing techniques. Masters thesis, Multimedia University.

Full text not available from this repository.
Official URL: http://erep.mmu.edu.my/

Abstract

Many of us are lucky to be able to see the world and go about our daily activities. Around the globe and in every part of the world, there are Visually impaired persons (VIPs), and they comprise a significant portion of the population. The visually impaired person may often face difficulty in reading banknotes. Malaysian researchers have proposed Ringgit banknote recognition systems for the visually impaired, but they only recognise banknote value, lacking counterfeit detection. Further detecting fake banknote is still a problem waiting to be solved by the banknote recognition system developers to improve accessibility and recognition for visually impaired users. This research discusses Malaysian banknote counterfeit detection algorithms that make use of image processing techniques and fuzzy logic algorithms. Image processing is a fast-growing technology that has impacted many businesses and industries. Furthermore, this research proposed two novel fuzzy logic algorithms for detecting counterfeit Malaysian banknotes: Fuzzy Logic based Weighted Averaging algorithm (FLWA) and Fuzzy Logic based Weighted Specific algorithm (FLWS). The FLWA and FLWS Malaysian Banknote Counterfeit Detection algorithms were intercompared with other parallel methods (MobileNet model using RMSprop Loss Function (learning_rate=0.0001) and VGG16 model using 2D Convolution Layer (32 neural) at TensorFlow's Keras API). In terms of accuracy, processing speed, and complexity, FLWA and FLWS outperform the TWO parallel methods shown in the experimental results. In addition, the FLWA and FLWS Malaysian Banknote Counterfeit Detection algorithms are intra-compared. However, FLWS’s accuracy is much higher than FLWA's in detecting Malaysian Counterfeit Banknotes. The number of security features is restricted to nine security elements. Banknotes were extracted and encoded using image processing techniques. Several security features, such as the magnifier and feeling mechanism, require a high-quality tool or an extension tool to be extracted. The Ringgit Counterfeit detection system was tested at the Malaysian Association for the Blind (MAB) and received positive feedback from participants.

Item Type: Thesis (Masters)
Additional Information: Call No.: TA1637 .T87 2023
Uncontrolled Keywords: Image processing—Digital techniques
Subjects: T Technology > TA Engineering (General). Civil engineering (General) > TA1501-1820 Applied optics. Photonics
Divisions: Faculty of Engineering and Technology (FET)
Depositing User: Ms Nurul Iqtiani Ahmad
Date Deposited: 28 Aug 2024 09:13
Last Modified: 28 Aug 2024 09:13
URII: http://shdl.mmu.edu.my/id/eprint/12873

Downloads

Downloads per month over past year

View ItemEdit (login required)