Perhitungan Frekuensi Kedipan Mata Berbasis Convolutional Neural Network

Daryanto, Atyantagratia Vidyasmara (2021) Perhitungan Frekuensi Kedipan Mata Berbasis Convolutional Neural Network. Undergraduate thesis, Institut Teknologi Sepuluh Nopember.

Text 07211640000032-Undergraduate_Thesis.pdf - Accepted Version Restricted to Repository staff only Download (12MB) | Request a copy

Abstract

Mata merupakan organ yang sangat sensitif, benda dan tekanan dengan ukuran kecil sudah dapat menimbulkan gangguan pada organ ini. Kedipan mata terjadi karena dipen- garuhi oleh sensitifitas kornea dan kekeringan mata. Terdapat dua parameter utama pada kedipan mata, yaitu frekuensi dan durasi, frekuensi berkedip dapat bertingkat salah satunya adalah saat orang dalam kondisi mengantuk. Pada peneli- tian ini dilakukan pendenteksian pada kondisi mata berkedip dari dataset video. Peneletian ini menggunakan disiplin ilmu Deep Learning dengan metode Convolutional Neural Network. Dataset berupa video akan diekstrak menjadi frame gambar, yang kemudian dilakukan labelling pada kondisi mata terbuka dan mata tertutup. Setelah mendapatkan hasil klasifikasi untuk mata terbuka dan mata tertutup, dibuat suatu model untuk mendeteksi kedipan mata pada video. Pada klasifikasi kondisi mata terbuka dan mata tertutup didapatkan akurasi sebesar 96% dan akan dilakukan pengujian pada video untuk deteksi kedipan mata secara otomatis.
===========================================================================================================
Eyes are very sensitive organ, objects and pressure
with a small size can cause interference with this organ. The blinking occurs because it is affected by the sensitivity of the cornea and the dryness of the eye. There are two main parameters in eye blinking, namely frequency and duration, the frequency of blinking can be graded, one of which is when people are sleepy. In this study, the detection of eye blinking conditions from the video dataset was carried out. This research uses the Deep Learning discipline with the Convolutional Neural Network method. The video dataset will be extracted into an image frame, which is then labeled with the eyes open and closed. After getting the classification results for open and closed eyes, a model is made to detect eye blinking on the video. In the classification of open and closed eyes, an accuracy of 96% is obtained and a video test will be carried out for automatic blink detection.

Item Type:	Thesis (Undergraduate)
Uncontrolled Keywords:	Kedipan Mata, Video, Klasifikasi, Convoltuional Neural Network, Eye-Blink, Videos, Classification, Convolutional Neural Network
Subjects:	T Technology > T Technology (General) T Technology > T Technology (General) > T57.5 Data Processing T Technology > TA Engineering (General). Civil engineering (General) > TA1637 Image processing--Digital techniques
Divisions:	Faculty of Electrical Technology > Computer Engineering > 90243-(S1) Undergraduate Thesis
Depositing User:	Atyantagratia Vidyasmara Daryanto
Date Deposited:	12 Mar 2021 05:40
Last Modified:	12 Mar 2021 05:40
URI:	http://repository.its.ac.id/id/eprint/84146

Actions (login required)

View Item