Malaysian Sign Language Translator for Mobile Application

Abstract

The society often faces a communication barrier with the deaf and mute community due to the lack of sign language translators. Hand gestures are used as a sign language to communicate between the ordinary people and the deaf people. However, the existing way to learn sign language is ineffective and inconvenient. The number of mobile translator applications for the Malaysian Sign Language on the market is less. The proposed system contains a sign detector mechanism using Support Vector Machine (SVM) to detect and interpret Malaysian Sign Language. The software used in this project were IntelliJ IDEA and Android Studio. This project's development consists of four main phases: dataset acquisition, dataset trainer, shape classification, and sign recognition. Each character is tested ten times to measure the performance of the developed system. The application is performed using an Android smartphone and successfully tested with character A to Z in a normal light condition with a maximum distance of 0.7 m. The successfully testing rate of this developed system is 90%. This application allows people to easily communicate with the deaf community because it is user-friendly and does not require an internet connection.