A comprehensive assistive solution for visually impaired persons

People with vision loss need to make their lives more independent. The proposed system is capable to provide multi solutions in a single package with the ability to switch their algorithm from online to offline. The proposed work provides the indoor Navigation system with Object detection, object recognition and distance depth calculation. The system is also able to generate audio for object positions and object recognition using Text to Speech technology (TTS), and also take audio input from users to find out the objects using Speech Recognition technology. The proposed system consists of a Processing unit, Camera (depth Camera Kinect V2) for object detection and recognition using the YOLO V3 (You Only Look Once) Algorithm, headphones for audio input from a user and audio output from the system and power supply in case of free movement it would be batteries. The camera is mounted on the chest (can be mounted on the head), so whenever the VIP wants to search for anything the system guides them about the position of the object, and if VIP wants to move, the system assists in avoiding obstacles as well as wayfinding in the form of audio. The average accuracy of performance is 93%.