Voice-assisted Smart Guidance System for Indoor Navigation for the Visually Impaired

To enhance the independence of visually impaired individuals in indoor spaces, this thesis introduces an iOS application tailored for the visually impaired, leveraging artificial intelligence and computer vision techniques. The primary research objectives are to enhance autonomy for the visually impaired through indoor navigation using only a mobile device, to evaluate the efficacy of machine learning algorithms for this task, and to assess performance on the iOS platform utilizing Apple's Neural Engine. The application's workflow involves receiving user input about the destination, identifying the location via the smartphone's camera, and estimating the distance to the destination. It integrates LiDAR, monocular depth estimation, and YOLOv8 object detection to determine navigable areas and to avoid obstacles. Auditory instructions then communicate navigation guidance. A key advantage of the system is its adaptability to different environments without the need for pre-mapping, thereby removing financial and hardware burdens. The methods optimize computational and energy efficiency. Moreover, the iOS platform provides performance benefits through hardware acceleration. The research questions examine the effectiveness of monocular depth estimation and YOLOv8 in aiding navigation, the advantages of the iOS platform and Apple's Neural Engine, and the scalability across various indoor environments. The literature review analyzes challenges specific to visually impaired individuals regarding indoor navigation and existing solutions. The methodology outlines the development process for the iOS application and the machine learning components. The discussion evaluates the system's performance and the utility of the machine learning algorithms. This thesis represents a significant advancement in assistive technology for the visually impaired by enabling autonomous indoor navigation using only a smartphone. It has considerable potential to improve autonomy and quality of life.