Pengenalan Ekspresi Wajah menggunakan Metode Enhanced Local Binary Patterns

Samosir, Yolanda Wisdanita (2019) Pengenalan Ekspresi Wajah menggunakan Metode Enhanced Local Binary Patterns. Other thesis, Institut Teknologi Sepuluh Nopember.

Text 05111540000023-Undergraduate_Thesis.pdf Restricted to Repository staff only Download (3MB) | Request a copy

Abstract

Facial Expression Recognition (FER) adalah salah satu bidang penelitian tentang pengenalan ekspresi wajah manusia secara komputasional. Pengenalan ekspresi wajah memainkan peran signifikan dalam interaksi manusia-komputer serta berbagai bidang ilmu perilaku dan menjadi salah satu topik penting dalam visi komputer. Pengenalan ekspresi wajah manusia mengkategorikan ekspresi wajah menjadi satu dari banyak kelas ekspresi wajah berdasarkan fitur gambar tersebut. Pada tugas akhir ini, pengenalan ekspresi wajah menggunakan data pelatihan dan uji coba yang diambil dari dataset “The Japanese Female Facial Expression” (JAFFE) dan “Karolinska Directed Emotional Faces” (KDEF) yang berisi foto wajah manusia dengan 7 ekspresi berbeda. Praproses terhadap data antara lain dilakukan proses perbaikan citra menggunakan Contrast Limited Adaptive Histogram Equalization (CLAHE), proses dekteksi bagian wajah menggunakan Haar Cascade, perubahan resolusi gambar menjadi 64x64 piksel pada citra wajah. Lalu pada hasil praproses data dilakukan ekstraksi fitur menggunakan Local Binary Patterns. Pada fitur LBP akan dilakukan seleksi fitur dengan memilih fitur di mana piksel LBP varians tinggi dipilih untuk mewakili wajah atau disebut Enhanced Local Binary Patterns. Dari metode Enhanced Local Binary Patterns didapatan nilai akurasi terbaik sebesar 79.64% dengan menggunakan classifier K-Nearest Neighbor (KNN) pada dataset KDEF.
==================================================================================================================================
Facial Expression Recognition (FER) is one of the fields of research on computational research on human facial expressions. Facial expression, a non-verbal communication, is a means through which humans convey their inner emotional state, thus playing an important role in social interaction. Facial expression recognition plays a significant role in human-computer interaction as well as various fields of behavioral science and is one of the important topics in computer vision. Human facial expression recognition categorizes an image of facial expression into one of many facial expression classes based on the features extracted from the image. In this undergraduate thesis, facial expression recognition using the train and test data taken from The Japanese Female Facial Expression (JAFFE) and “Karolinska Directed Emotional Faces” (KDEF) dataset which contains human facial expression with 7 different expressions. The preprocessing of the images include applying Contrast Limited Adaptive Histogram Equalization (Clahe) to improve the image contrast, applying face detection using Haar Cascade, changing the image resolution to 64x64 pixels on face digital images . Then after preprocessing, feature extraction is done using Local Binary Patterns. Then a feature selection will be applied by selecting the feature where the high variant LBP pixels to represent faces or later called Enhanced Local Binary Patterns. This method performed the best accuracy value of 79.64% by using the K-Nearest Neighbor as a classifier on the KDEF dataset.

Item Type:	Thesis (Other)
Uncontrolled Keywords:	Contrast Limited Adaptive Histogram Equalization, Dataset Karolinska Directed Emotional Faces, Dataset The Japanese Female Facial Expression, Enhanced Local Binary Patterns, K-Nearest Neighbor, Pengenalan Ekspresi Wajah Manusia
Subjects:	T Technology > T Technology (General) T Technology > T Technology (General) > T57.5 Data Processing T Technology > T Technology (General) > T58.5 Information technology. IT--Auditing
Divisions:	Faculty of Information and Communication Technology > Informatics > 55201-(S1) Undergraduate Thesis
Depositing User:	Yolanda Wisdanita Samosir
Date Deposited:	23 Jul 2026 03:38
Last Modified:	23 Jul 2026 03:38
URI:	http://repository.its.ac.id/id/eprint/67023

Actions (login required)

View Item