Journal
IDENTIFIKASI POLA SIDIK JARI BERBASIS TRANSFORMASI WAVELET DAN JARINGAN SYARAF TIRUAN PROPAGASI BALIK
Unnes Physics Journal · Vol. 4 No. 1 (2015)
- ISSN
- 2252-6978
- Published
- 2015-08-26
- Publisher
- Unnes Physics Journal
- Language
- eng
Abstract
Sistem identifikasi dengan memanfaatkan karakteristik biometrik saat ini telah berkembang secara luar biasa dan mampu menggantikan sistem identifikasi secara konvensional. Salah satu karakteristik biometrik manusia adalah sidik jari. Suatu pola sidik jari normal terdiri dari ridge dan vallays yang dapat dianalisis pada tingkat lokal (menggunakan minutiae) dan global (non-minutiae). Pada penelitian ini, identifikasi dibangun berdasarkan pendekatan global (non-minutiae) yaitu menggunakan transformasi wavelet sebagai pengolah awal (preprocessing) dan jaringan syaraf tiruan propagasi balik sebagai pengambil keputusannya (metrika). Metode pengambilan citra sidik jari dilakukan secara inkless menggunakan sensor sidik jari dan pemprosesan citra dilakukan dengan menggunakan software MATLAB. Hasil penelitian menunjukkan bahwa semua jenis wavelet memiliki kinerja yang baik dalam menghasilkan ciri. Hal ini ditunjukan dengan nilai Euclidean yang kecil ketika ciri dari citra sampel yang sama diuji. Jarak Euclidean terkecil dihasilkan oleh wavelet haar, yaitu sebesar 1,9. Hasil performa jaringan menunjukan bahwa tingkat keakuratan menurun dengan bertambah besar basis data yang digunakan dan meningkat dengan bertambah kecil nilai MSE jaringan. Nilai keakuratan tertinggi didapat sebesar 96.67 % untuk basis data latih 15 pola sidik jari (30 citra pembanding) dengan MSE jaringan sebesar 10-6. Identification system by employing biometric characteristics recently has grown tremendously and is able to replace the conventional identification systems. One of the characteristics of human beings is a fingerprint biometric. A normal fingerprint pattern consisting of ridge and vallays which can be analyzed on the local level (using minutiae) and global (non-minutiae). In this study, the identification is built based on a global approach (non-minutiae) that uses wavelet transformation as initial processing (preprocessing) and backpropagation neural network as decision makers (Metrika). Fingerprint image retrieval method performed inkless using the fingerprint sensor and the image processing is done by using MATLAB software. The results showed that all types of wavelet has good performance in generating characteristics. It is indicated by small euclidean value when the characteristic of the image of the same sample tested. The smallest euclidean distance generated by haar wavelet of 1.9. Network performance results showed that the level of accuracy decreases with increase in number of data base and increases with decreases in value of MSE of network. The highest value of accuracy that is obtained is 96.67% when the training data base is 15 fingerprint patterns (30 images comparator) and the value of MSE of network is 10-6.
Collaboration network of the authors
36 authors · 109 co-authorship links
Hover or tap a name to highlight its co-authors. Double-click to open the author's page.
- State University of Semarang
- Institut Teknologi Sepuluh Nopember
- Universitas Airlangga
- Affiliation unknown
- Different organisations
- Same organisation
- Affiliation unknown
- Thicker line = more joint articles · bigger circle = more articles
Show as table
| Co-author | Organisation | Joint articles |
|---|---|---|
| Sukiswo Supeni Edi | Affiliation unknown | 2 |
| Abdul Latif | Affiliation unknown | 1 |
| Sri Wulan Siti Khotijah | Affiliation unknown | 1 |
| Sri Endah Ardhi Ningrum Abdullah | Affiliation unknown | 1 |
| Dwi Atmoko | Affiliation unknown | 1 |
| Achmad Sopyan | Affiliation unknown | 1 |
| Imam Ghozali | Institut Teknologi Sepuluh Nopember | 1 |
| Ani Rusilowati | State University of Semarang | 1 |
| Arneta Dwi Safitrtri | Affiliation unknown | 1 |
| Khumaedi | State University of Semarang | 1 |
| Khisni Rohendy | Affiliation unknown | 1 |
| Achmad Rifa’i | Affiliation unknown | 1 |
| Teguh Febrianto | Affiliation unknown | 1 |
| - Susilo | Affiliation unknown | 1 |
| Retno Endah Savitri | Affiliation unknown | 1 |
| Vera Rizchi Cahyani Putri | Affiliation unknown | 1 |
| sukiswo Supeni Edie | Affiliation unknown | 1 |
| - Rohmad | Affiliation unknown | 1 |
| Langlang Handayani | Affiliation unknown | 1 |
| Teguh Darsono | Affiliation unknown | 1 |
| Slamet Haryono | Affiliation unknown | 1 |
| Haydnn Caesha Maulana | Affiliation unknown | 1 |
| R. Indriyanto | Affiliation unknown | 1 |
| Indrawan Nur Cahyono | Affiliation unknown | 1 |
| Aditya Bagus Wicaksana | Affiliation unknown | 1 |
| Khusnul Ain | Universitas Airlangga | 1 |
| Amun Amri | Affiliation unknown | 1 |
| Zhong Tao Jiang | Affiliation unknown | 1 |
| Yola Bertilsya Hendri | Affiliation unknown | 1 |
| Yoyok Dwi Setyo Pambudi | Affiliation unknown | 1 |
| Mazhibayev Assylzhan | Affiliation unknown | 1 |
| Kambarova Elmira | Affiliation unknown | 1 |
| Khairulazhar Jumbri | Affiliation unknown | 1 |
| Chun-Chen Yang | Affiliation unknown | 1 |
Affiliations come from OpenAlex; journal OAI-PMH metadata carries none. Authors without an OpenAlex ID are matched by name, which can merge or split people.