The problem Drishti solves

Our project intends to solve a very basic problem faced by visually impaired people i.e; navigating through obstacles. As they don't have any visual input from surroundings, so they cannot take any non-intuitive decision to avoid the obstacles and this usually leads to accidents. So in order to solve this problem we are providing them Computer-Aided Navigation system which can analyse visual input , take decisions and aid them in navigation by sending electrical impulses directly to the muscles using electrodes. Solving this problem is important as visually impaired people find a lot of difficulty in navigating through obstacles especially in a new unknown environment.They generally use cane (which are not that much effective) or else they need to depend on human assistance. As we are using visual analysis of obstacles it can be done quite effectively using Computer-Vision and its efficiency can be increased manifold using Machine Learning. We are using quite simple electronic equipments like Raspberry PI, Pi camera module, Muscle stimulator ,Muscle Stimulator electrodes etc. As the equipments are not very complex it is easy to manufacture and repair the product and due to this fact only the price of the product is quite low and is compatible to a smartphone.

Challenges we ran into

Lack of Precison in Equipments(link ultrasonic Sensor). Integration of realtime video processing in Raspberry pi 3 B(Because of 1GB of RAM). Biological knowledge of Saritorial muscle. Lot of indentation errors.