【摘 要】
:
为了解决高清视频的畸变校正及显示的实时性问题,提出了一种CUDA架构下的并行加速方案。系统利用张正友标定方法获得摄像机的内部参数和畸变参数,并利用GPU的大规模并行计算能力加速校正过程。校正后,位于显存的图像数据直接利用OPENGL驱动进行显示。针对不同架构GPU片上资源限制不同,设计了一种并行划分参数自整定算法,保证了程序移植到不同GPU后能充分利用硬件资源,实现最佳性能。实验结果表明,本文设计
论文部分内容阅读
为了解决高清视频的畸变校正及显示的实时性问题,提出了一种CUDA架构下的并行加速方案。系统利用张正友标定方法获得摄像机的内部参数和畸变参数,并利用GPU的大规模并行计算能力加速校正过程。校正后,位于显存的图像数据直接利用OPENGL驱动进行显示。针对不同架构GPU片上资源限制不同,设计了一种并行划分参数自整定算法,保证了程序移植到不同GPU后能充分利用硬件资源,实现最佳性能。实验结果表明,本文设计的系统对传统串行处理系统的综合加速比最高可达39倍以上,对2 596×1 920分辨率视频下的处理帧率可
其他文献
The rapid growth of Internet usage has dramatically changed the way we interact with the outside world.Many people read news,communicate with friends and purchase goods online.These activities are usu
Efficient management of task flow plays an important role in prediction,planning,control and optimization of tasks.In this paper,according to major indicators to evaluate during task flow management s
Efficient video summarization leads to facilely exploring video content appropriate to the users intention with low cognitive demand.In this paper,we present a novel approach for summarizing videos in
Fisheye view is an effective approach to visualizing and navigating large data sets by offering both local details and global context in the same view.However,by using different magnification factors
Thematic information of a long document(i.e.,a novel) can be multi-faceted:an interleaving of multiple topics,a sequential evolution of a set of themes,or a crossing superimposition of topics and them
Real estate is an important industry in most countries.However,the analysis of the real estate market is very challenging as the data are high dimensional and have complex spatial and temporal pattern
This paper proposes a VLSI architecture of the optimized Speeded-Up Robust Feature (SURF) algorithm. The SURF algorithm which is widely used in computer vision applications, locates interest points (I
Image-based visual servoing can be used to efficiently control the motion of robot manipulators. When the initial and the desired configurations are distant, however, as pointed out by many researcher
Automatic recognition of the contents of a scene is an important issue in the field of computer vision. Although considerable progress has been made, the complexity of scenes remains an important chal
空间时间延迟积分电荷耦合器件(TDICCD)相机图像压缩2维(2-D)离散小波变换模块的基于常用硬件实现方法实现效率低,为此提出一种适于电荷耦合器件(CCD)图像的(128,N)快速2-D提升小波实现结构.在CCD图像行方向上,采用16个提出的1维(1-D)行提升小波变换模块快速实现结构并行分解图像,每个模块采用4个延时寄存器的值作为中间值并用2步预测更新步骤融合的结构来计算行小波系数;在行小波系