Design And Development Of Assistive Mobile Application For Visually Impaired Using Machine Learning

Mohamad Napiah, Nur Aliah (2019) Design And Development Of Assistive Mobile Application For Visually Impaired Using Machine Learning. Project Report. UTeM, Melaka, Malaysia. (Submitted)

Text Design And Development Of Assistive Mobile Application For Visually Impaired Using Machine Learning.pdf Download (312kB)

Abstract

This project is about the development of assistive mobile application for visually impaired people using machine learning. The project is mainly focused on visually impaired people as they face problems of social mobility in their daily routine. Even though there are several of assistive technology tools for visually impaired people, their needs have not yet been fulfilled. Accordingly, an assistive mobile application based on machine learning is developed to help the user with sound output with the name of the object when the camera captures the object’s image. Objects focused in this project are money, cloths and basic things of visually impaired use in their daily basis. This project works by capturing real-time images and the application is contained TensorFlow Object Application Programming Interface (API) that uses Single Shot Detector (SSD) with a pre-trained model trained using MobileNet V2 model developed at Google dataset. The captured image will be compared to the preloaded image dataset to determine the project output. Objects that have been detected with a score greater than 0.5 will be displayed. The application interface will provide object names and confidence scores.

Actions (login required)

View Item