Evaluating Deep Learning and Traditional Approaches Used in Source Camera Identification

TL;DR

This paper compares deep learning and traditional methods like PRNU and JPEG analysis for source camera identification, evaluating their accuracy and discussing future development needs for real-world application.

Contribution

It provides a comparative analysis of CNNs, PRNU, and JPEG artifacts in source camera identification, highlighting their strengths and limitations.

Findings

CNNs achieve higher accuracy than traditional methods.

PRNU remains effective for certain device classifications.

JPEG artifact analysis offers complementary information.

Abstract

One of the most important tasks in computer vision is identifying the device using which the image was taken, useful for facilitating further comprehensive analysis of the image. This paper presents comparative analysis of three techniques used in source camera identification (SCI): Photo Response Non-Uniformity (PRNU), JPEG compression artifact analysis, and convolutional neural networks (CNNs). It evaluates each method in terms of device classification accuracy. Furthermore, the research discusses the possible scientific development needed for the implementation of the methods in real-life scenarios.