Showing papers in "Journal of Visual Communication and Image Representation in 2016"

TL;DR: A new database of aerial images provided as a tool to benchmark automatic target recognition algorithms in unconstrained environments and gives the performance of baseline algorithms on this dataset, for different settings of these algorithms, to illustrate the difficulties of the task and provide baseline comparisons.

TL;DR: In this article, a mathematical model called the Rectified-Correlations on a Sphere (RECOS) is proposed to answer two fundamental questions about the structure of the convolutional neural networks (CNN): (1) why a nonlinear activation function is essential at the filter output of all intermediate layers.

TL;DR: An extendable multi-cues integration framework for face anti-spoofing using a hierarchical neural network is proposed, which can fuse image quality cues and motion cues for liveness detection.

TL;DR: A comprehensive review of the recent progress in image segmentation, covering 190 publications, gives an overview of broad segmentation topics including not only the classic unsupervised methods, but also the recent weakly-/semi- supervised methods and the fully-super supervised methods.

TL;DR: An in-depth survey of the existing image mosaicing algorithms by classifying them into several groups, and the fundamental concepts are first explained and then the modifications made to the basic concepts by different researchers are explained.

TL;DR: In this article, the authors propose a Plug-and-Play-Prior approach for coupling Gaussian denoising algorithms to Poisson noisy inverse problems, which is based on a general approach termed "Plug-and Play-Prior".

TL;DR: The efficacy of the proposed deep learning approach to estimate the number of individuals presented in a mid-level or high-level crowd visible in a single image is demonstrated by quantifying the counting performance through multiple evaluation criteria.

TL;DR: The experimental results demonstrate that the proposed DCT based effective self-recoverable fragile watermarking scheme not only outperforms high-quality restoration effectively, but also removes the blocking artifacts and improves the accuracy of tamper localization due to the use of very small size blocks, smoothing function and two levels tampering detection mechanisms.

TL;DR: A fast reference frame selection algorithm for the HEVC encoder that efficiently removes the encoding complexity of the best reference frame decision process and the rate distortion performance degradation is negligible.

TL;DR: The PVO embedding is extended into a 2D form by integrating the pairwise prediction- error expansion, and a reversible 2D mapping adapted to the special distribution of prediction-error pairs is proposed.

TL;DR: An improved pixel-value-ordering based reversible data hiding based on multiple histograms modification (MHM) is proposed to achieve a high-performance RDH and experimental results show the superiority of the new algorithm over the state-of-the-art methods.

TL;DR: A novel deep-sea imaging method that compensates for the attenuation discrepancy along the propagation path, and an effective underwater scene enhancement scheme are described, characterized by a reduced noise level, better exposure of dark regions, and improved global contrast such that the finest details and edges are significantly enhanced.

TL;DR: A robust 2D Otsu’s thresholding method that improves the robustness to Salt&Pepper noise and Gaussian noise significantly and introduces a region post-processing step to deal with the pixels of noise and edges.

TL;DR: A deformable part-based correlation filter tracking approach which depends on coupled interactions between a global filter and several part filters, which gives significantly better results compared with the state-of-the-art trackers.

TL;DR: Quantitative and qualitative results obtained from experiments on a wide variety of natural scene images demonstrate the effectiveness of the proposed approach over other methods at reducing artefact while increasing image contrast and colourfulness.

TL;DR: Properties of quaternion Exponent moments (QEMs) are discussed in detail in detail and a robust color image zero-watermarking algorithm is proposed which is robust to geometric attacks.

TL;DR: A new ALPD method is proposed which effectively detects LP area from an image in the hazardous conditions, and is tested on 850 car images having different hazardous conditions.

TL;DR: A single image deblurring algorithm to remove spatially variant defocus blur based on the estimated blur map and adopts a BM3D-based non-blind deconvolution algorithm to restore the latent image.

TL;DR: A novel method of multimodal medical image fusion using fuzzy-transform (FTR) based fusion helps in preservation as well as effective transfer of detailed information present in input images into a fused image.

TL;DR: Two new reversible data hiding algorithms are proposed for the encrypted images that are more applicable in cloud computing without degrading the security level and can be losslessly transmitted in the different applications of Paillier cryptosystem.

TL;DR: This work addresses the issue of drops in visual quality of recent video watermarking methods by introducing a robust and imperceptible non-blind color video frame water marking algorithm.

TL;DR: The experimental results show that the proposed scheme is effective, especially for the multi-objects splicing, and is proven to be superior to the existing state-of-the-art method.

TL;DR: Experimental results show that the proposed approach outperforms state of the art methods for underwater fish detection and a Poisson-Gauss mixture algorithm is proposed for noise reduction.

TL;DR: A novel ROI-based reversible data hiding scheme in encrypted medical images that can be losslessly recovered after extracting the embedded data.

TL;DR: This study shows that gait feature can effectively distinguish from different human beings through a novel representation - relative distance-based gait features and indicates that the relative distance feature is quite effective and worthy of further study in more general scenarios.

TL;DR: Experimental results demonstrate that, MTSD significantly outperforms texton co-occurrence matrix, multi-texton histogram, micro-structure descriptor and saliency structure histogram.

TL;DR: In this paper, an iterative nonlocal means filter (INLM) is proposed to exploit the image non-local similarity feature in the S&P noise removal procedure, and the proposed iterative framework update the similarity weights and the estimated values for higher accuracy.

TL;DR: A robust watermarking framework with a blind extraction process for HEVC encoded video, where a readable watermark is embedded invisibly in intra predicted blocks of the HEVC encode video, enforces security by exploring the spatio-temporal characteristics of the compressed video and a random key for selection of embedding regions.

TL;DR: This paper proposes a novel multiplicative contourlet domain watermark detector based on using the Maximum Likelihood (ML) decision rule and BKF distribution and demonstrates the high efficiency of Bessel K form (BKF) distribution to model these coefficients.

TL;DR: The proposed detection scheme outperforms some state-of-the-art methods when applied to Columbia Image Splicing Detection Evaluation Dataset (DVMM), and ranks fourth in phase 1 on the Live Ranking of the first Image Forensics Challenge.