Kerangka Kerja Berbagi Data Gambar dengan Fitur Keterlacakan Data Menggunakan Skema Watermarking dan Integrasi Blockchain

Putra, I Kadek Agus Ariesta (2025) Kerangka Kerja Berbagi Data Gambar dengan Fitur Keterlacakan Data Menggunakan Skema Watermarking dan Integrasi Blockchain. Masters thesis, Institut Teknologi Sepuluh Nopember.

Text 6025231078-Master_Thesis.pdf - Accepted Version Restricted to Repository staff only Download (20MB) | Request a copy

Abstract

Penelitian ini mengusulkan skema berbagi data gambar menggunakan metode watermarking citra reversibel yang terintegrasi dengan sistem pelacakan berbasis blockchain, dengan tujuan utama untuk menjaga integritas, keaslian, dan jejak distribusi data digital. Evaluasi eksperimen dilakukan terhadap sepuluh citra berwarna dengan resolusi 512×512 piksel dari domain umum, medis, dan biometrik, di antaranya citra umum dari USC-SIPI; citra medis dari CHASEDB1. KSAVIR, dan ISIC; serta citra biometrik dari FERET. Metode dievaluasi menggunakan metrik kuantitatif peak signal-to-noise ratio (PSNR), structural similiarty index measure (SSIM), dan normalized correlation (NC), serta diuji terhadap sepuluh jenis serangan umum dan tiga serangan adversarial. Hasil menunjukkan bahwa metode yang diusulkan secara konsisten menghasilkan nilai PSNR di atas 50 dB dan SSIM mendekati 1. Nilai NC tetap tinggi (≥ 0,95) pada sebagian besar skenario serangan, mengindikasikan ketahanan watermark terhadap distorsi. Sistem ini juga menunjukkan akurasi pelacakan sumber distribusi sebesar 100% pada seluruh kombinasi citra dan serangan, melampaui dua metode pembanding. Selain itu, efisiensi waktu penyematan dan ekstraksi masing-masing tercatat sebesar 0,0836 detik dan 0,0176 detik. Implementasi smart contract pada jaringan Polygon Mainnet menunjukkan rata-rata konsumsi gas sebesar 48.380 unit per transaksi, sedangkan sistem IPFS menunjukkan ketersediaan 100% dengan latensi pengambilan data rata-rata 0,815 detik. Secara keseluruhan, metode ini menawarkan keseimbangan antara imperseptibilitas, ketahanan, dan kemampuan pelacakan, serta layak diterapkan dalam sistem distribusi data digital.
====================================================================================================================================
This study proposes an image data sharing scheme using a reversible image watermarking method integrated with a blockchain-based tracking system, with the primary goal of preserving the integrity, authenticity, and distribution traceability of digital data. Experimental evaluations were conducted on ten color images with a resolution of 512×512 pixels from general, medical, and biometric domains, including general images from USC-SIPI; medical images from CHASEDB1, KSAVIR, and ISIC; and biometric images from FERET. The method was assessed using quantitative metrics: Peak Signal-to-Noise Ratio (PSNR), Structural Similarity Index Measure (SSIM), and Normalized Correlation (NC), and was tested against ten types of common attacks and three adversarial attacks. The results show that the proposed method consistently produces PSNR values above 50 dB and SSIM values close to 1.00. The NC values remain high (≥ 0.95) in most attack scenarios, indicating strong watermark robustness against distortion. The system also achieved 100% accuracy in tracing the source of distribution across all image and attack combinations, outperforming two baseline methods. In addition, the embedding and extraction processes demonstrated high efficiency, with average execution times of 0.0836 seconds and 0.0176 seconds, respectively. The implementation of the smart contract on the Polygon Mainnet showed an average gas consumption of 48,380 units per transaction, while the IPFS system achieved 100% availability with an average data retrieval latency of 0.815 seconds. Overall, the proposed method offers a well-balanced combination of imperceptibility, robustness, and traceability, making it suitable for deployment in digital data distribution systems.

Item Type:	Thesis (Masters)
Uncontrolled Keywords:	Blockchain, Data Sharing, Data Tracking, Reversible Image Watermarking
Subjects:	T Technology > T Technology (General) > T201 Patents. Trademarks T Technology > T Technology (General) > T57.5 Data Processing T Technology > T Technology (General) > T58.5 Information technology. IT--Auditing
Divisions:	Faculty of Intelligent Electrical and Informatics Technology (ELECTICS) > Informatics Engineering > 55101-(S2) Master Thesis
Depositing User:	I Kadek Agus Ariesta Putra
Date Deposited:	25 Jul 2025 04:14
Last Modified:	25 Jul 2025 04:14
URI:	http://repository.its.ac.id/id/eprint/121561

Actions (login required)

View Item