Implementation Of Hand Recognition Using Convolutional Neural Network For Sign Language

Rahman, Rian and , Aris Rakhmadi, S.T., M. Eng. (2023) Implementation Of Hand Recognition Using Convolutional Neural Network For Sign Language. Skripsi thesis, Universitas Muhammadiyah Surakarta.

	PDF (Naskah Publikasi) Naspub_L200164014_Rian Rahman-3.pdf Download (818kB)
	PDF (Surat Pernyataan Publikasi) Publikasi.pdf Restricted to Repository staff only Download (475kB)

Abstract

The deaf and mute are hard to communication by talking. They using sign language to communication towards the other and the general public. Sign language using hand to communication. Indonesia has two sign languages, namely Bahasa Isyarat Indonesia (BISINDO) and Sistem Isyarat Indonesia (SIBI). Technological development made a dictionary to translating a hand gesture. Machine learning is the technology that make a computer can recognize a hand from the image. Using hand gesture recognition to read gesture of a hand, that is a technology that can define gesture of human hand using mathematical algorithms. Resolution of the dataset can affect a percentages of success rates. Convolution Neural Network (CNN) is a method of Deep Learning that used in this study. Using CNN to hand gesture recognition can make the peed and the accuracy become higher. This study about implementation of CNN method to 24 alphabets of SIBI as the input in image. The input data is 3468 images to be trained. The test data consists of 720 images, each alphabet consisting of 30 images. Testing with data that has been trained produces an accuracy of 96,67%.

Item Type:	Thesis (Skripsi)
Uncontrolled Keywords:	CNN, deep learning, hand recognition, SIBI, sign language.
Subjects:	Q Science > QA Mathematics > QA75 Electronic computers. Computer science > QA751 Artificial Intelligence Q Science > QA Mathematics > QA75 Electronic computers. Computer science > QA753 Computer Systems and Networks
Divisions:	Fakultas Ilmu Komunikasi dan Informatika > Teknik Informatika
Depositing User:	RIAN RAHMAN
Date Deposited:	21 Feb 2023 08:45
Last Modified:	21 Feb 2023 08:45
URI:	http://eprints.ums.ac.id/id/eprint/110235

Actions (login required)

View Item