Enkripsi Citra Digital dengan IDEA Cipher: Pengacakan Blok Pseudorandom LCG dan Evaluasi Melalui Image Correlation Analysis

Abstract

The use of cryptography methods in digital image processing has become a crucial focus of research in
safeguarding visual information. This study analyzes the utilization of the Linear Congruential Generator (LCG)
as a block scrambling tool in the image encryption process using the International Data Encryption Algorithm
(IDEA). The aim of this research is to evaluate the effectiveness of LCG in reducing correlation patterns among
pixel blocks in encrypted images. In experiments conducted using image datasets from OpenDataset, the test
results indicated a significant difference between encrypting images using regular IDEA and IDEA enhanced with
LCG. The block scrambling method using LCG successfully generated encrypted images with correlation
coefficient values approaching zero, signifying its effectiveness in obscuring correlation patterns among pixel
blocks. Despite a decrease in Mean Squared Error (MSE) and slight degradation in the quality of the encrypted
images, the evaluation results affirmed a significant improvement in the average correlation coefficient by 4.57.
This indicates an enhanced security level for encrypted images against cryptanalysis leveraging correlation
patterns, while considering compromises in image quality. The findings of this research affirm that utilizing LCG
in block scrambling during image encryption with IDEA can produce encrypted images with higher security levels
by reducing correlation patterns, albeit with a slight sacrifice in image quality. The implications of these findings
might contribute to the development of security methods in digital image processing in the future