Design of Sign Language Recognition Using E-CNN

Abstract

Most people do not understand sign language, so they need a bridge for the community to be able to communicate with deaf people. Technology that continues to develop and continues to strive to help humans, can be a solution that can be used to create a communication bridge between the community and deaf people, the use of technology that can be used is the use of image processing technology as a translator tool. Image processing can translate images into text. In the implementation of digital image processing, it will use the hand key point library, where the hand key point library is a library that will detect the location of the hand in each image, but as it is known, image processing cannot stand alone as a data processor but requires an algorithm that functions as a classification tool. The Convolutional Neural Network (CNN) algorithm in the Deep Learning method can be a classification tool, with the ability of the Convolutional Neural Network (CNN) to learn several things. And according to several previous studies that combining several algorithms can increase the accuracy value. In this study, a trial of combining CNN models using the Ensemble method has been successfully carried out with the results being able to increase the accuracy value to 99.4%. So that the results of the research can be summarized that using Ensemble can increase the higher accuracy value.