AI Powered Assistive Solution for Navigation of Visually Impaired Persons

This paper explores the development of smart navigation stick for the visually impaired. The innovation provides crucial audio feedback and precise object location information, significantly improving the overall security of blind users. The specific functionalities of each component and the speculation for the choices are specified. The arrangement that has been recommended takes processing power of the Raspberry Pi 3 Model B to check the Raspberry Pi camera V2. The camera captures real-time images for object detection and recognition. Furthermore, an ultrasonic sensor (HC-SR04) is incorporated To determine how far away obstacles are, providing a complementary layer of spatial awareness. To enhance object recognition accuracy, the author utilizes the SSD MobileNet algorithm alongside OpenCV libraries. After detecting objects, they are converted to audio feedback for the user via the Python eSpeak module. This multi-modal approach caters to diverse user preferences and environmental conditions.