Image blur is a useful source of information for many applications. Shallow depths of field (DoF), where non-subject parts of the image are heavily blurred, is a signature element in professional photography and film editing. DoF effects are also known to improve photorealism, to mediate monocular depth perception, and to make a focused object attract attention. An important component in image manipulation tools is the ability to fake blur to hide splicing and copy-move operations.;This dissertation focus on exploring blur properties for light field depth reconstruction and photo forensic.;Depth recovery from focus/defocus: we introduce a bilateral consistency metric on the surface camera (SCam) for depth from light field refocusing to handle significant occlusions. The concept of SCam is used to model angular radiance distribution with respect to a 3D point. Our bilateral consistency metric is used to indicate the probability of occlusions by analyzing the SCams. We further show how to distinguish between infocus and defocus, textured and non-textured, and Lambertian and specular through bilateral SCam analysis. To speed up the matching process, we apply the edge-preserving guided filter on the consistency-disparity curves. Experimental results show that our technique outperforms both the state-of-the-art and the recent light field stereo matching methods, especially near occlusion boundaries.;Robust focal stack symmetry: we describe a technique to recover depth from a light field (LF) using two proposed features of the LF focal stack. One feature is the property that non-occluding pixels exhibit symmetry along the focal depth dimension centered at the in-focus slice. The other is a data consistency measure based on analysis-by-synthesis, i.e., the difference between the synthesized focal stack given the hypothesized depth map and that from the LF. These terms are used in an iterative optimization framework to extract scene depth. Experimental results on real Lytro and Raytrix data demonstrate that our technique outperforms state-of-the-art solutions and is significantly more robust to noise and undersampling.;Image splicing detection from blur analysis: we present a new technique based on the analysis of the camera response functions (CRF) for efficient and robust splicing and copy-move forgery detection and localization. We first analyze how non-linear CRFs affect edges in terms of the intensity-gradient bivariate histograms. We show distinguishable shape differences between real and forged blurs near edges after a splicing operation. Based on our analysis, we introduce a deep-learning framework to detect and localize forged edges. In particular, we show the problem can be transformed to a handwriting recognition problem and resolved by using a convolutional neural network. We generate a large dataset of forged images produced by splicing followed by retouching and comprehensive experiments show our proposed method outperforms the state-of-the-art techniques in accuracy and robustness.;Focus manipulation detection: we look to detect focus manipulations which may obscure important details in a photo, and develop an algorithm which handles manipulations where the blur is approximately consistent with the scene geometry. Such manipulations are easily generated, for instance, by modern smartphone cameras having multiple imagers to infer depth, e.g. 'Portrait Mode' of the iPhone7Plus. Our algorithm incorporates multiple cues, including edge anomalies, noise variance, JPEG artifacts, and demosaicing statistics, to discern manipulated images from optically blurred images with 97% classification accuracy.
展开▼