Legowo, Rafiidha Selyna (2017) Klasifikasi Gerakan Tangan SIBI (Sistem Isyarat Bahasa Indonesia) Menggunakan Leap Motion dengan Metode Klasifikasi Naive Bayes. Undergraduate thesis, Institut Teknologi Sepuluh Nopember.
Preview |
Text
2913100028-Undergraduate_Thesis.pdf - Published Version Download (10MB) | Preview |
Abstract
Pada penelitian tugas akhir ini dilakuan pembuatan sistem pelatihan isyarat Bahasa Indonesia menggunakan alat Leap motion. Sistem tersebut mampu mengenali gerakan isyarat bahasa dan dapat digunakan sebagai metode pelatihan isyarat bahasa bagi masyarakat awam sehingga dapat mengatasi halangan dalam hal berkomunikasi antara penyandang disabilitas tuli dan orang normal yang ingin mempelajari gerakan isyarat bahasa. Dalam penelitian ini, pengambilan data dilakukan menggunakan alat Leap Motion. Terpilih sebanyak 10 gerakan isyarat bahasa yang terdiri atas lima gerakan statis dan pengembangan dinamisnya. Dilakukan proses ekstraksi ciri sehingga menghasilkan sebanyak 19 ciri. Algoritma klasifikasi yang digunakan dalam penelitian ini adalah menggunakan metode Naive Bayes. Semua data yang diolah dalam sistem mengacu pada kamus Sistem Isyarat Bahasa Indonesia (SIBI) yang sudah resmi diakui oleh pemerintah dan melambangkan kosa kata dan abjad Bahasa Indonesia. Setelah dilakukan pengujian terhadap dua macam subjek, yang pertama adalah dengan subjek yang terdapat pada data training dihasilkan nilai akurasi rata-rata sebesar 80.5%, sedangkan pengujian terhadap subjek yang tidak terdapat dalam data training mengalami penurunan nilai akurasi yakni hanya mencapai nilai rata-rata 70.7%.
============================================================
In this final project, Indonesian sign language training system using Leap motion has been made. The system is able to identify the movement of sign languages and can be used as a method of sign language training to the common people so as to overcome barriers in terms of communicating between deaf people and normal people. In this research, data acquisition is done by using Leap Motion. There are 10 gestures chosen as the research object consists of five static gestures and its dynamic development. The feature extraction process was performed to produce a total of 19 features. The classification algorithm used in this research is using the Naive Bayes method. All data processed in the system refers to the Sistem Isyarat Bahasa Indonesia (SIBI) dictionary that has been officially recognized by the government and symbolizes Indonesian vocabulary and alphabet. After testing the training data obtained an accuracy of 80.5% in experiments in the ideal environment with the same subject as in training data and relatively lower accuracy on the experiments with user whom is not available in training data, the application only get 70.7% accuracy rate.
| Item Type: | Thesis (Undergraduate) |
|---|---|
| Additional Information: | RSKom 004.019 Leg k |
| Uncontrolled Keywords: | Leap Motion, SIBI, Naive Bayes |
| Subjects: | L Education > LC Special aspects of education T Technology > TK Electrical engineering. Electronics Nuclear engineering > TK5105.546 Computer algorithms |
| Divisions: | Faculty of Information Technology > Computer Engineering > 90243-(S1) Undergraduate Thesis |
| Depositing User: | Rafiidha Selyna Legowo |
| Date Deposited: | 12 Oct 2017 03:43 |
| Last Modified: | 08 Mar 2019 02:23 |
| URI: | http://repository.its.ac.id/id/eprint/46150 |
Actions (login required)
 |
View Item |