Blind source camera identification

Abstract

An interesting prohlem in digital forensics is that given a digital image, would it he possihle to identify the camera model which was used to obtain the image. In this paper we look at a simplified version of this problem by trying to distinguish between images captured by a limited number of camera models. We proposc i number of features which could be used by a classifier to idcntify the source camera of an image in a blind manner. We also provide experimental results and show reasonable accuracy in distinguishing images from the two and five different camera models using the proposed features.

Source Camera Identification Using Support Vector Machines

“…Each image was divided into 512 × 512 non-overlapping blocks of which eight were randomly chosen * 2.0 * * * * 5.7 * 20.2 * * * 1.2 * 69. 8 for analysis. The image database consisted of 4,600 different images with 512 × 512 resolution.…”

“…Several researchers have employed multi-class classifiers for camera identification [1,4,8,11,14]. The methodology involves extracting feature vectors from several image samples created by various camera models.…”

“…Although this method has an average classification accuracy of nearly 92% for six different cameras, it cannot easily distinguish between cameras of the same brand but different models. The classification accuracy can be improved by combining the feature vector in [8] with the lens radial distortion coefficients of digital cameras [4]. However, extracting distorted line segments to estimate the distortion parameters limits the application of this method to images that contain distorted line segments.…”

“…These methods primarily extract features from the digital image and use a classifier to determine image origin. The method of Kharrazi, et al [8] uses image color characteristics, image quality metrics and the mean of wavelet coefficients as features for classification. Although this method has an average classification accuracy of nearly 92% for six different cameras, it cannot easily distinguish between cameras of the same brand but different models.…”

Scene Illumination as an Indicator of Image Manipulation

“…[37,33].Some of the existing approaches are classification based [6,22]. Unfortunately, the outcome of these algorithms is often hard to interpret for nontechnical surveyors, e.g.…”