<![CDATA[Digital image manipulation has become increasingly difficult to identify through visual inspection, creating a need for digital forensic methods capable of objectively verifying image authenticity. This study aims to analyze the characteristics and compare the performance of Error Level Analysis (ELA), Noise Analysis, and Clone Detection in detecting image splicing and copy-move forgery. The study used 500 digital images, consisting of 250 original images and 250 manipulated images, which were analyzed using the three forensic methods through a Python-based application and evaluated based on their detection success rates. The results show that ELA achieved detection rates of 70.40% for image splicing and 64.00% for copy-move forgery, while Noise Analysis achieved 39.20% and 28.00%, respectively, and Clone Detection achieved 46.40% and 81.60%. These findings indicate that ELA is more effective for detecting compression-based manipulation, Clone Detection performs better in identifying copy-move forgery, whereas Noise Analysis serves as a complementary method for analyzing inconsistencies in noise patterns.]]>
Copyrights © 2026
