Evaluasi Kualitas Gambar CCTV Pada Kinerja Pengenalan Wajah

Pratama, Rifqi (2019) Evaluasi Kualitas Gambar CCTV Pada Kinerja Pengenalan Wajah. Other thesis, Institut Teknologi Sepuluh Nopember.

[thumbnail of 07111340000185-Undergraduate_Thesis.pdf]
Preview
Text
07111340000185-Undergraduate_Thesis.pdf

Download (2MB) | Preview

Abstract

Kinerja sistem pengenalan wajah seperti yang diaplikasikan di Smart Surveillance System pada bandara erat kaitannya dengan kualitas citra wajah. Kendala yang umumnya dialami sistem pengenalan wajah adalah soal masalah parameter kualitas citra seperti pencahayaan, blur, dan noise yang terdapat pada citra sehingga kinerja sistem tidak bisa maksimal.
Penggunaan eigenface dalam sistem pengenal wajah digunakan untuk mereduksi dimensi citra wajah, sehingga menghasilkan variabel yang lebih sedikit yang memudahkan citra untuk diobservasi. Dengan mencocokkan citra yang diuji terhadap citra latih berdasarkan nilai eigenface. Pencocokan didasari pada perhitungan jarak euclidean terdekat.
Pengujian sistem pengenalan wajah menggunakan eigenface dengan menguji citra dengan parameter kualitas citra yang bervariasi menyimpulkan bahwa prosentase akurasi keberhasilan sistem pengenalan wajah adalah 77,78%.
============================================================================================================================
The facial recognition system performance as applied at Smart Surveillance System at the airport is closely related to image quality of the face. Common imagery problems that affected performances of the system are about image quality parameter such as illumination, blur, and noise in the images so the system isn’t able to work optimally.
Using eigenface in facial recognition system to reduce images dimension so that it generates less variable numbers thus the images are easier to be observed. By matching test images toward train images. Image matching based on nearest euclidean distance calculation.
The facial recognition system using eigenface method by testing images with various image quality parameters, the percentage of successful face recognition system is 77,78%.

Item Type: Thesis (Other)
Uncontrolled Keywords: face recognition, eigenface, euclidean distance, pengenalan wajah, eigenface, jarak euclidean.
Subjects: T Technology > TK Electrical engineering. Electronics Nuclear engineering > TK7882.P3 Pattern recognition systems
Divisions: Faculty of Electrical Technology > Electrical Engineering > 20201-(S1) Undergraduate Thesis
Depositing User: Rifqi Pratama
Date Deposited: 21 Sep 2020 03:03
Last Modified: 12 Jan 2024 07:32
URI: http://repository.its.ac.id/id/eprint/81989

Actions (login required)

View Item View Item