Object Detection and Identification for Visually Impaired

Visual impairment takes a major toll in a person's life for them to act independently without others help. This paper introduces an assistive technology that makes use of the YOLO (You Only Look Once) algorithm integrated into wearable spectacles along with a camera linked to a Raspberry Pi processor. The system aims to perform real-time object detection and identification, transforming visual data into audio feedback for users. The hardware setup involves a camera connected to a Raspberry Pi for data processing captured using the camera. An earphone is connected to the Raspberry Pi to guide and navigate the user. The YOLO algorithm, specifically optimized for real-time usage on the Raspberry Pi, distinguishes objects captured within the camera's view. Consequently, the system translates this visual data into understandable audio information transmitted to users via connected earphones by using the trained data. This paper introduces a methodology, execution, and outcomes of the proposed system. It offers a thorough insight into the hardware configuration, the algorithmic framework employed for object detection, and the integration methodology for generating audio-based feedback. Evaluations of the system's performance is calculated by its efficacy, swiftness, and precision in real-world situations, highlighting its potential to aid visually impaired individuals in understanding unfamiliar environments.