REAL TIME OBJECT DETECTION MOBILE APPLICATION INTERFACE DEVELOPMENT FOR LOW VISION

• Main Author: Gillian Tekad Tuerah, Felicia
• Format: Final Project
• Language: Indonesia
• Online Access: https://digilib.itb.ac.id/gdl/view/69177
• Institution: Institut Teknologi Bandung

Based on the BPS (Badan Pusat Statistik) survey in 2020, there are 22.5 million people with disabilities in Indonesia, which is about 5% of the total population of Indonesia. Accessibility for persons with disabilities is regulated in UU No. 8 of 2016 which is in the form of facilities provided for persons with disabilities in order to realize equal opportunities. Low vision is one type of disability which will then be the focus of this final project. Low vision is a state of partial vision loss that cannot be corrected by prescription glasses, contact lenses, or surgery. Therefore, the problem that is often faced by people with low vision is the difficulty in identifying objects around them. This final project aims to solve the problems faced by people with low vision in order to increase the productivity of their daily life. The solution that developed in the implementation of this final project is a mobile application for detecting objects in real time for people with low vision. The application has main features, namely object detection in real time and several other supporting features in the form of assistance features, flash features, and mode change features. The application is developed with a native development approach on the Android platform. The development of the application interface is based on the results of the high-fidelity prototype design and the flow of the application interface. The results of the application interface implementation were tested in two tests, namely user interface testing and user acceptance testing. The results of user interface testing show that the application is in accordance with the design generated on the high-fidelity prototype and is in accordance with the application interface flow. The results of user acceptance testing conducted on 5 participants with low vision showed that 5 out of 5 participants successfully completed 11 of 11 testing tasks so that the application was in accordance with user needs and could interact well with users.