Showing papers in "Journal of Visual Communication and Image Representation in 2014"

TL;DR: A new representation of human actions called Sequence of the Most Informative Joints (SMIJ), which is extremely easy to interpret and performs better than several state-of-the-art algorithms for the task of human action recognition.

TL;DR: An effective method to recognize human actions using 3D skeleton joints recovered from 3D depth data of RGBD cameras is proposed and a new action feature descriptor, EigenJoints, is designed which combine action information including static posture, motion property, and overall dynamics.

TL;DR: Multi-resolution surfel maps as compact RGB-D image representation that integrates color and depth measurements seamlessly in a multi-resolution map representation for simultaneous localization and mapping and is used in mobile manipulation tasks.

TL;DR: This work provides a means for multimodal gait recognition, by introducing the freely available TUM Gait from Audio, Image and Depth (GAID) database, which simultaneously contains RGB video, depth and audio.

TL;DR: This paper proposes an encrypted signal-based reversible data hiding (ESRDH) with public key cryptosystem, not only for images, and the proposed scheme is secure based on Paillier homomorphic encryption.

TL;DR: Experimental result shows that the proposed scheme significantly outperforms the previous approaches on reversible data hiding in encrypted images based on lossless compression of encrypted data.

TL;DR: This article presents an information gain-based variant of the next best view problem for occluded environment that utilizes a belief model of the unobserved space to estimate the expected information gain of each possible viewpoint.

TL;DR: A new framework to detect and recognize stairs, pedestrian crosswalks, and traffic signals based on RGB-D (Red, Green, Blue, and Depth) images is developed.

TL;DR: A customizable human kinematic model that extracts skeletons from RGB-D sensor data that adapts on-line to difficult unstructured scenes taken from a moving camera and benefits from using both color and depth data is presented.

TL;DR: This paper deals with shape extraction from depth images (point clouds) in the context of modern robotic vision systems by presenting various optimizations of the 3D Hough Transform used for plane extraction from point cloud data.

TL;DR: This paper proposed a new SVD-based image watermarking by embedding the principal component of a watermark into the host image of block based manner using spread spectrum concept and demonstrates that the proposed method overcomes the false positive problem, achieves a high payload, and outperforms the former reliable SVD.

TL;DR: An improved DBC method based on the original one for improvement of the accuracy can solve the two kinds of problems which the DBC has: over-counting boxes along z direction and under- Counting boxes just at the border of two neighboring box blocks where there is a sharp gray-level abruption exits.

TL;DR: Experimental results confirm the merits of the proposed algorithm in providing the intra-frame error propagation-free advantage, the quality improvement for marked images, the compression power inherited from HEVC, and the superiority of embedding capacity for low bitrate coding when compared with the previous two algorithms for H.264/AVC.

TL;DR: A framework for pose-based human action recognition is proposed, where coordinates of joints are considered as inputs and vectors of dissimilarities to a set of prototype actions are computed.

TL;DR: A PCA-based dictionary building for sparse representation and classification of universal facial expressions and the framework outperforms the recognition rate of the state-of-the-art techniques by about 6%.

TL;DR: A novel framework for color image retrieval through combination of all the low level features, which gives higher retrieval accuracy is presented, which improves the average retrieval accuracy by approximately 16% and 14% over CDH and ART respectively.

TL;DR: This work proposes, for the best of the author's knowledge, the first publicly available RGB-D benchmark dataset with hand-labeled ground truth of several challenging scenarios to test background/foreground segmentation algorithms.

TL;DR: A novel objective full-reference measure of image quality (VPSNR), which is a modified PSNR measure that for images compressed by block-based compression algorithms (like JPEG) the proposed measure in the pixel domain matches well with MOS.

TL;DR: The MVC+D standard, which specifies an MVC-compatible MVD coding format, is reviewed, which enables depth-image-based rendering (DIBR) of additional viewpoints in the decoding side and hence helps in such advanced application scenarios.

TL;DR: A novel K-means variant that takes into account the periodic nature of angular data is applied on the former and Nearest neighbour and SVM classification are used for action recognition on the test data.

TL;DR: A model to calculate the average coding efficiency based on subjective quality scores, i.e., mean opinion scores (MOS) based on fItted Curves (SCENIC), which is expected to report more realistic coding efficiency.

TL;DR: Experimental results have shown the suitability of the proposed approach for tamper detection and accurate authentication and the designed criteria were used to judge the received image by classifying it into: authenticated, incidentally or maliciously attacked.

TL;DR: A partial volume reconstruction of the frontal surface of each silhouette is done and a novel feature termed as Pose Depth Volume (PDV) is derived from this volumetric model.

TL;DR: Segmentation of unknown objects in cluttered scenes using Grahp-Cut for segmentation of raw RGB-D data into parametric surface patches and learning of perceptual grouping between surfaces with SVMs.

TL;DR: The Cayley representation of spatial lines is developed, which is a nonlinear minimal parametrization circumventing the tiresome Plucker constraint and derives a novel line projection function which outperforms the previous works both in efficiency and accuracy.

TL;DR: A framework for recognizing the sign language alphabet based on the combination of three shape descriptors based on a set of geometric features derived from the convex hull that encloses the hand shape taking into account the hand orientation.

TL;DR: One of the transformations, RDgDb, which requires just 2 integer subtractions per image pixel, on average results in the best ratios for JPEG2000 and JPEG XR, while for a specific set or in case of JPEG-LS its compression ratios are either the best or within 0.1 bpp from the best.

TL;DR: An approach which performs scene parsing and data fusion for a 3D-LIDAR scanner (Velodyne HDL-64E) and a video camera and the fused results are more reliable than that acquired via analysis of only images or Velodyne data.

TL;DR: The proposed system combines the efficiency of the Bag-of-visual-Words strategy and the power of graphs for structural representation of features followed by a graph-based video representation which models the spatio-temporal relations among these features.

TL;DR: In this article, a calibration procedure for depth and color cameras was proposed to improve the accuracy of the 3D measurements of a RGB-D camera and to co-register different calibrated devices.