Recognizing Medication Labels System with Audio Output for Individuals with Visual Impairments

Managing medication poses a significant challenge for individuals with visual impairments as they struggle to access essential information on packaging. To overcome this obstacle, a computer vision-based system has been developed to assist visually impaired users in independently recognizing medication details, with a particular emphasis on drug strip packaging. This innovative system integrates advanced technologies for text detection and recognition. Initially, the YOLOv8 algorithm is employed for text detection, achieving a reported training accuracy of 91.84%. Once detected, easyOCR performs Optical Character Recognition (OCR) to extract the text. The identified information is then converted into speech using the pyttsx3 text-to-speech synthesis module, providing auditory output for the user. The system has been developed and deployed on a laptop, ensuring reliable performance and accessibility for users. It successfully completes each stage, from initial text detection to final speech output, with an average total response time of approximately 4.5 seconds. Given these performance metrics, this solution holds great promise in enhancing medication management independence for visually impaired individuals and reducing the risk of medication errors.