AI Model Integrated Omani Currency Detector for the Visually Impaired-A Comparative Analysis
S. Sangeetha, Asan Banu Jinnah, Annamalai Senthil Kumar, T. Porselvi, Abrar Abdulmohsin Salim Alajmi, Bayan Abdulmohsin Salim Alajmi, Aisha Issa Ali Alzadjali, Aisha Taaeeb Alhamdani · 2024
In this research work a mobile application is created specifically for the blind and individuals with visual impairmentsfor enabling them to recognize Omani currenciesin real time independently. Also,people those who have a significant reduction in their ability to see, and it impacts their daily life usesvoice commandto recognize the Omani cash currencies. The proposed currency detector works by taking a picture of Omani currency by voice command, then the cash identification is done by using a teachable machine and personal image classifier AI models, at lastthe phone tells the value of currency correctly as a voice massage to the user, whether it is one riyal or five riyals or ten riyals or twenty riyals. This application can be installed both in IOS and Android devices. The proposed system aims to accurately identify and classify different denominations of Omani Rial (OMR) notes in real-time, enabling visually impaired individuals to confidently handle monetary transactions without any assistance. The application consists of two screens. Screen 1 set for teachable machine classifier AI model and Screen 2 as personal image classifier AI model through which a picture of the Omani currency is taken, identify the currencies and the user knows the cash value independently without others help by sound. The proposed workis tested by the teachable machine AI model for ten times and identified the Omani currencies correctlyfor 7 times,also tested by personal image classifier AI model for ten times and got the correct cash result for 8 times. From the test results,it is found that the personal image classifier works more accurate and better than teachable machine AI model.