A Comprehensive Study on Supermarket Indoor Navigation for Visually Impaired using Computer Vision Techniques

Abstract

The ability to navigate is a fundamental skill that every person must have. People who are visually challenged need regular support when travelling from one place to another. It is difficult for them to have an autonomous shopping experience in a supermarket. The project’s goal is to make one of the daily life’s task easier for visually impaired persons. The project entails aiding visually impaired customers by guiding them to the destination of the corresponding product sections using an indoor navigation technique. The program aids in providing the visually impaired with shopping environment as well as assistance on how to buy their preferred goods. When logging into the Android app, the user or someone who is blind will speak about their shopping list preferences (voice input). The camera is turned on to provide the live-streaming video. The user can then begin navigating indoors to reach the section where the product is found. This incorporates the technology that is used in automated self-driving cars. Modeled highdefinition simulated maps of the supermarkets are used to navigate inside employing the shortest distance. The first item from the list is selected, and the user’s location is identified by matching the features in the image frame with the map and used to compute the path to the product section from there. It provides speech output for navigation. Someone who assists the people to get the products from shelves will help them to get the exact item. The buying process continues until the last item on the list is bought. As a result, this will make tasks comparatively easier for visually impaired persons and make indoor navigation possible.