Effect of Error Level Analysis on The Image Forgery Detection Using Deep Learning

Abstract

Digital image modification or image forgery is easy to do today. The authenticity verification of an image become important to protect the image integrity so that the image is not being misused. Error Level Analysis (ELA) can be used to detect the modification in image by lowering the quality of image and comparing the error level. The use of deep learning approach is a state-of-the-art in solving cases of image data classification. This study wants to know the effect of adding ELA extraction process in the image forgery detection using deep learning approach. The Convolutional Neural Network (CNN), which is a deep learning method, is used as a method to do the image forgery detection. The impacts of applying different ELA compression levels, such as 10, 50, and 90 percent, were also compared in this study. According to the results, adopting the ELA feature increases validation accuracy by about 2.7% and give the better test accuracy. However, the use of ELA will slow down the processing time by about 5.6%.

Mass Classification of Breast Cancer Using CNN and Faster R-CNN Model Comparison

“…The difference between deep learning and machine learning is in data management performance. Deep learning can manage an increasing amount of extensive data and solve a problem, while machine learning optimally processes small amounts of data [20] [21]. The use of deep learning in this study can adapt to updating data and recognizing disease functional and structural characteristics [22].…”

WCBnet: Weighted Convolutional Block Modelling of Signed-Value Error Levels for Image-Wise Copy-Move and Splicing Detection

“…A dense layer and flatten layer are used as a classifier for ELA results of forged face images [11]. Input images were pre-processed by ELA and classified by a simple two-layer or three-layer convolutional neural network, regarding splicing and copy-move respectively [12]. ELA was applied on UNet to achieve pixel-level image forgery detection, and improved the pixel-level F1-score to 0.686 [13].…”

Neural Network Based Data Encryption: A Comparison Study among DES, AES, and HE Techniques

“…2) Attacks on encrypted image: In this research, ELA image analysis [45], lightwave adjustment, and Zsteg analysis have been applied to encrypted images to test the resistance of encrypted images (Figure 6). In Photo-Forensics's ELA analysis (Error Level Analysis), the JPEG quality is set to 90%, the error scale is set to 20, and the opacity value is specified as 0.52, indicating the degree of transparency of an element in the image.…”