DESIGN OF IDENTIFICATION, VOICE INTERFACE AND UART COMMUNICATION ON SYSTEM FOR IDENTIFYING SILENT OBJECTS FOR PERSONS WITH BLINDNESS ON PUBLIC ROADS

• Main Author: Made Iwan Darmawan, I
• Format: Final Project
• Language: Indonesia
• Online Access: https://digilib.itb.ac.id/gdl/view/43113
• Institution: Institut Teknologi Bandung

More than one percent or around 3.5 million Indonesians experience blindness with Java as the region with the highest number of blind people. The existence of visual limitations causes blind people to need conventional aids and other people in carrying out activities. The main problem now is that conventional tools have limitations in identifying straightforward objects. The System for Identifying Silent Objects for Persons with Blindness on Public Roads is a system designed to overcome these problems. This system is designed to be able to identify dangerous objects that are in front of users such as trees, poles, trenches and crossings. The system can take pictures and identify them through computational processes . Some important subsystems owned by this system such as identification, voice interface and serial communication (UART). The identification subsystem has the function of processing image to identify objects. Voice interface will provide voice information to the user about object in front of him. Meanwhile, the UART communication subsystem has function to send position data that has been processed by Raspberry Pi to Arduino for mapping the position of objects. Based on the model used, the SSD Mobilenet has a weakness to identifying objects that have pixels size below 128x90 pixels The result show that, the system is able to identify object like trees, poles, trenches and crossings with accuracy for the identified object at 80,5% and provide the information about location and type of objects.