Rashmi Rameshrao Shrirao <br/><br/>
AI-Driven Drug Pill Recognition System: A CNN-Based Android Application for Visually Impaired and Senior Citizens 

 - pp24-33 
<br/><br/> As individuals age, challenges such as declining vision and memory can increase the risk of medication errors, particularly among the elderly and visually impaired. To address this issue, this research presents a deep learning-based Android application for accurate and accessible drug pill recognition. The system leverages a contrast-enhanced Convolutional Neural Network (CNN) trained on a diverse pill image dataset, achieving a test accuracy of 98%. Integrated with a REST API, the model enables real-time image classification via a smartphone camera. The application further enhances usability through voice-assisted feedback and visual pill details, promoting autonomy and medication adherence. This AI-driven solution bridges the gap between healthcare and technology, offering a practical tool to reduce medication errors and improve the quality of life for users with visual and cognitive impairments. 
<br/><br/><label>Subjects--Facted Index Terms: </label><br/>Pill Recognition<br/>Deep Learning<br/>Elderly Healthcare<br/>Convolutional Neural Network (CNN)<br/> Assistive Technology.