Current Issue : July - September Volume : 2012 Issue Number : 3 Articles : 5 Articles
Slow-motion replays are content full segments of broadcast soccer videos. In this paper, we propose an efficient method for\r\ndetection of slow-motion shots produced by high-speed cameras in soccer broadcasts. A rich set of color, motion, and cinematic\r\nfeatures are extracted from compressed video by partial decoding of the MPEG-1 bitstream. Then, slow-motion shots are modeled\r\nby SVM classifiers for each shot class. A set of six full-match soccer games is used for training and evaluation of the proposed\r\nmethod. Our algorithm presents satisfactory results along with high speed for slow-motion detection in soccer videos....
Block-based connected components labeling is by far the fastest algorithm to label the connected components in\r\n2D binary images, especially when the image size is quite large. This algorithm produces a decision tree that\r\ncontains 211 leaf nodes with 14 levels for the depth of a tree and an average depth of 1.5923. This article attempts\r\nto provide a faster method for connected components labeling. We propose two new scan masks for connected\r\ncomponents labeling, namely, the pixel-based scan mask and the block-based scan mask. In the final stage, the\r\nblock-based scan mask is transformed to a near-optimal decision tree. We conducted comparative experiments\r\nusing different sources of images for examining the performance of the proposed method against the existing\r\nmethods. We also performed an average tree depth analysis and tree balance analysis to consolidate the\r\nperformance improvement over the existing methods. Most significantly, the proposed method produces a\r\ndecision tree containing 86 leaf nodes with 12 levels for the depth of a tree and an average depth of 1.4593,\r\nresulting in faster execution time, especially when the foreground density is equal to or greater than the\r\nbackground density of the images....
To prepare images for better segmentation, we need preprocessing applications, such as smoothing, to reduce\r\nnoise. In this paper, we present an enhanced computation method for smoothing 2D object in binary case. Unlike\r\nexisting approaches, proposed method provides a parallel computation and better memory management, while\r\npreserving the topology (number of connected components) of the original image by using homotopic\r\ntransformations defined in the framework of digital topology. We introduce an adapted parallelization strategy\r\ncalled split, distribute and merge (SDM) strategy which allows efficient parallelization of a large class of topological\r\noperators. To achieve a good speedup and better memory allocation, we cared about task scheduling and\r\nmanaging. Distributed work during smoothing process is done by a variable number of threads. Tests on 2D\r\ngrayscale image (512*512), using shared memory parallel machine (SMPM) with 8 CPU cores (2Ã?â?? Xeon E5405\r\nrunning at frequency of 2 GHz), showed an enhancement of 5.2 with cache success rate of 70%....
Spectral images as well as color images observed from object surfaces are much influenced by various illumination\r\nconditions such as shading and specular highlight. Several invariant representations were proposed for these\r\nconditions using the standard dichromatic reflection model of dielectric materials. However, these representations\r\nare inadequate for other materials like metal. This article proposes an invariant representation that is derived from\r\nthe standard dichromatic reflection model for dielectric and the extended dichromatic reflection model for metal.\r\nWe show that a normalized surface-spectral reflectance by the minimum reflectance is invariant to highlights,\r\nshading, surface geometry, and illumination intensity. Here the illumination spectrum and the spectral sensitivity\r\nfunctions of the imaging system are measured in a separate way. As an application of the proposed invariant\r\nrepresentation, a segmentation algorithm based on the proposed representation is presented for effectively\r\nsegmenting spectral images of natural scenes and bare circuit boards....
Acoustic simulation has always played an important role in the development of new ultrasound imaging\r\ntechniques. In nonlinear ultrasound imaging particularly, the simulators are accurate but time-consuming, because\r\nof the high derivative order of the propagation equation and to the classic solution based on finite difference\r\nschemes. This article presents a fast 3D + t nonlinear ultrasound simulator, based on a generalized angular\r\nspectrum method, particularly fit for the graphics processing unit (GPU). Indeed, the Fourier domain approach\r\ndecreases the derivative order of the propagation, thus significantly speeding up the simulation time. The simulator\r\nwas implemented and optimized on a central processing unit (CPU) and a GPU, respectively. The processing times\r\nmeasured on two different graphic cards show that, compared to the CPU, GPU-based implementation is 3.5-13.6\r\ntimes faster....
Loading....