Pengenalan Sidik Jari Berminyak Menggunakan Histogram of Oriented Gradients (HOG) dan Artificial Neural Network (ANN)".

Baihaqi, Ahmad Raihan (2025) Pengenalan Sidik Jari Berminyak Menggunakan Histogram of Oriented Gradients (HOG) dan Artificial Neural Network (ANN)". Other thesis, Institut Teknologi Sepuluh Nopember.

Text 2040211095-Undergraduate _Thesis.pdf Restricted to Repository staff only Download (9MB) | Request a copy

Abstract

Mesin filler pada PT. Charoen Pokphand Indonesia dilengkapi dengan sistem verifikasi sidik jari yang berfungsi untuk pencatatan riwayat petugas yang bertanggung jawab pada area filler. Biometrik sidik jari digunakan untuk autentikasi dan identifikasi karena memiliki minutiae (percabangan) yang bersifat unik dan permanen bagi setiap individu. Akan tetapi, petugas di lapangan sering kali gagal melakukan verifikasi akibat kondisi lingkungan yang ekstrem di area filler yang menyebabkan jari petugas berminyak sehingga mengakibatkan minutiae sulit dikenali dan terjadi kegagalan verifikasi oleh sensor. Untuk mengatasi permasalahan tersebut, penelitian ini menggunakan beberapa pendekatan preprocessing dan ekstraksi fitur menggunakan Histogram of Oriented Gradients (HOG). Nilai dari fitur sidik jari kemudian dilatih dan diidentifikasi menggunakan metode Artificial Neural Network (1764-512-256-10). Dataset mencakup 400 citra (10 subjek dengan 200 citra bersih + 200 citra berminyak) dengan 75% sebagai training dan 25% testing. Hasil menunjukkan akurasi pengenalan sebesar 98% untuk klasifikasi pada kondisi jari bersih dan berminyak. Serta akurasi identifikasi sebesar 80,8% menggunakan threshold EER 0,95 untuk penurunan jumlah False Accept dan False Reject. Peningkatan akurasi pengenalan sidik jari pada jari berminyak sebesar 94% dari 0%.
======================================================================================================================================
The filler machines at PT. Charoen Pokphand Indonesia are equipped with fingerprint verification system that records the history of personnel responsible for the filling area. Fingerprint biometrics are used for authentication and identification because they have unique and permanent minutiae (branching) for each individual. However, field personnel often fail to verify due to extreme environmental conditions in the filling area, which cause their fingers to become oily, making minutiae difficult to recognize and resulting in verification failure by the sensor. To address this issue, this study employs several preprocessing and feature extraction approaches using the Histogram of Oriented Gradients (HOG) method. The fingerprint feature values are then trained and identified using the Artificial Neural Network method (1764-512-256-10). The dataset includes 400 images (10 subjects with 200 clean images + 200 oily images), with 75% used for training and 25% for testing. The results show a recognition accuracy of 98% for classification under clean and oily finger conditions. Additionally, the identification accuracy is 80,8% using an EER threshold of 0,95 to reduce the number of False Accepts and False Rejects. The recognition accuracy of fingerprints on oily fingers improves to 94% from 0%.

Item Type:	Thesis (Other)
Uncontrolled Keywords:	Artificial Neural Network (ANN), Citra Sidik Jari, Fingerprint Image, Fingerprint Recognition, Histogram of Oriented Gradient (HOG), Pengenalan Sidik Jari.
Subjects:	T Technology > T Technology (General) > T57.5 Data Processing T Technology > TA Engineering (General). Civil engineering (General) > TA1637 Image processing--Digital techniques. Image analysis--Data processing. T Technology > TA Engineering (General). Civil engineering (General) > TA1650 Face recognition. Optical pattern recognition. T Technology > TK Electrical engineering. Electronics Nuclear engineering > TK7882.B56 Biometric identification T Technology > TK Electrical engineering. Electronics Nuclear engineering > TK7882.P3 Pattern recognition systems
Divisions:	Faculty of Vocational > 36304-Automation Electronic Engineering
Depositing User:	Ahmad Raihan Baihaqi
Date Deposited:	04 Aug 2025 09:30
Last Modified:	04 Aug 2025 09:30
URI:	http://repository.its.ac.id/id/eprint/125505

Actions (login required)

View Item