图像处理系统中的新型动态重构技术

阐述了一种新型动态结构可重构技术及其在分布式图像处理系统中的应用。以并行处理中的流水线技术和空间并行技术为基础，结合二者的优点提出一种可动态重构的结构模型，通过决策代价函数进行最优判决，将分解后算法任务最佳地映射到体系结构中去，在时间维和空间维上提高整个系统的实时并行处理性能。在实际环境中验证了该动态技术的正确性和优越性。     

利用串口快速传输图像并实时显示

针对图像处理系统调试中图像显示的问题,介绍了一种方便的图像观测方法,开发了一套串口通讯软件,实现了大量图像数据的串口快速传输与实时显示,并给出了部分源代码.     

Image Sequence Stabilization in Real Time

This paper describes a method ofstabilizingimage sequences obtained by a camera carried by a ground vehicle. The motion of the vehicle can usually be regarded as consisting of a desired smooth motion combined with an undesired non-smooth motion that includes impulsive or high-frequency components. The goal of the stabilization process is to correct the images so that they are approximately the same as the images that would have been obtained if the motion of the vehicle had been smooth.We analyse the smooth and non-smooth motions of a ground vehicle and show that only the rotational components of the non-smooth motion have significant perturbing effects on the images. We show how to identify image points at which rotational image flow is dominant, and how to use such points to estimate the vehicle's rotation. Finally, we describe an algorithm that fits smooth (ideally, piecewise constant) rotational motions to these estimates; the residual rotational motion can then be used to correct the images. We have obtained good results for several image sequences obtained from a camera carried by a ground vehicle moving across bumpy terrain.     

Fault Tolerance Using Dynamic Reconfiguration on the POEtic Tissue

Fault tolerance is a crucial operational aspect of biological systems and the self-repair capabilities of complex organisms far exceeds that of even the most advanced electronic devices. While many of the processes used by nature to achieve fault tolerance cannot easily be applied to silicon-based systems, in this paper we show that mechanisms loosely inspired by the operation of multicellular organisms can be transported to electronic systems to provide self-repair capabilities. Features such as dynamic routing, reconfiguration, and on-chip reprogramming can be invaluable for the realization of adaptive hardware systems and for the design of highly complex systems based on the kind of unreliable components that are likely to be introduced in the not-too-distant future. In this paper, we describe the implementation of fault tolerant features that address error detection and recovery through dynamic routing, reconfiguration, and on-chip reprogramming in a novel application specific integrated circuit. We take inspiration from three biological models: phylogenesis, ontogenesis, and epigenesis (hence the POE in POEtic). As in nature, our approach is based on a set of separate and complementary techniques that exploit the novel mechanisms provided by our device in the particular context of fault tolerance.     

基于GPU的实时图像拼接

大视野、高质量的图像信息对地面移动机器人的遥控操作具有非常重要的意义。提出了一种基于先验信息的自适应图像拼接方法。该方法在图像大致重叠区域中均匀选取待匹配点,利用改进的具有旋转不变性的NCC(Normalized Cross Correlation,归一化互相关)匹配方法进行区域相似性度量,通过RANSAC(Random Sample Con-sensus,随机采样一致性)算法估计图像射影变换模型,采用线性淡入淡出法进行图像融合。利用GPU强大的并行处理能力对算法进行了并行化实现,使图像拼接效率比单独采用CPU提高了60倍以上,稳定的拼接速度可达21.3fps。     

基于GPU的变型SIFT算子实时图像配准

针对SIFT变型算法描述向量维数过高实、时性差的问题,分别在建立高斯尺度金字塔、关键点的亚像素定位等方面进行改进与并行化。利用CUDA设备构架在GPU硬件上实现多线程,一方面避免了PCA方法造成的关键点信息流失,另一方面使得配准速度达到了工程中的实时性要求。在VS2005平台上通过C语言实现混合CUDA编程,结果表明该方法使得配准速度和点对匹配正确率都有较大提升。     

基于小波变换的实时影象匹配

讨论了一种基于小波变换的实时影像匹配方法。在多尺度特征提取的过程中，采用了局域正则性指数作为约束条件，保证了特征提取的可靠性；利用小波正交分解与重建实现了分层匹配；同时，采用特征与影像叠加进行了亚像素匹配。经过试验证明，上述方法是行之有效的。     

基于DSP的实时图像跟踪系统的优化设计

本文详细介绍了基于高性能TigerSHARC DSP处理模块和模板匹配算法(template matching)的实时图像跟踪处理系统的优化设计方法;深入分析了SAD操作中涉及到的地址对齐问题,提出了一种优化的设计方案,将并处理效率提高20倍.并在实际的实时图像跟踪系统中得到应用.     

基于FPGA与Sobel算法的实时图像处理

如今,图像处理算法的复杂度越来越高,图像处理的数据量越来越大,图像处理的实时性显得十分重要。为了解 决图像预处理、视频流数据实时性存在的问题,给出了一种基于FPGA和OV5640以Sobel算子进行边缘检测的图像采集与处 理系统设计方法,FPGA将OV5640摄像头采集到的视频流数据传送至SDRAM,由Sobel算子模板处理后通过VGA显示视频 图像。该设计基于Intel公司的Cyclone IV系列FPGA芯片EP4CE10F17C8进行了验证。实验结果表明,基于FPGA和Sobel边 缘检测算法,使用流水线设计和乒乓操作,可实现视频流数据处理的实时性。     

动态模拟红外背景图像应用

本文根据红外图像的特点和天空背景中各景物的红外辐射规律，提出和完成了动态红外背景图像的合成及其在应用中的模拟显示方法。     

一种全景图像的动态显示方法

为了实现动态实时显示全景图像的目的，文章提出了全景图像的合成及用Borland C Builder编程语言实现的方法，达到了预期的效果，结论证明该方法合理可行。     

基于深度纹理的实时碰撞检测算法

结合层次包围盒和基于图形硬件的方法,以带深度纹理的包围盒替代物体的几何模型,利用图形硬件在纹理映射时进行深度比较,以实现碰撞检测.实验结果表明,与CULLIDE算法相比,文中算法执行效率更高且执行时间固定,具有较高的实时性.     

Rendering Grass in Real Time with Dynamic Lighting

The authors present a real-time grass rendering technique that works for large, arbitrary terrains with dynamic lighting, shadows, and a good parallax effect. A novel combination of geometry and lit volume slices provides accurate, per-pixel lighting. A fast grass-density management scheme allows the rendering of arbitrarily shaped patches of grass.     

动态非均质半透明物体的实时绘制算法

现有的对半透明物体的绘制算法中大部分都是为均质材质设计的,虽然有少数算法可以绘制非均质材质,但效率不高,为此提出一种图形硬件加速的动态非均质半透明物体的实时绘制算法.首先将模型的采样点组织成八叉树的形式,将双向次表面散射反射函数(BSSRDF)表达绑定为顶点属性,同时充分利用图形硬件的并行性来实现八叉树的快速建立和遍历;然后以树的结点之间的空间距离和材质的相似度为准则提出一种新的采样方法来提高表面积分的效率.由于整个过程不需要任何预计算,因此该算法可以很容易地扩展到动画场景或者变化的材质,并且还可以灵活地选择BSSRDF的表达方式.实验结果表明,文中算法可以达到实时的绘制帧率和良好的视觉效果.     

Real Time Multiscale Rendering of Dense Dynamic Stackings

Dense dynamic aggregates of similar elements are frequent in natural phenomena and challenging to render under full real time constraints. The optimal representation to render them changes drastically depending on the distance at which they are observed, ranging from sets of detailed textured meshes for near views to point clouds for distant ones. Our multiscale representation use impostors to achieve the mid-range transition from mesh-based to point-based scales. To ensure a visual continuum, the impostor model should match as closely as possible the mesh on one side, and reduce to a single pixel response that equals point rendering on the other. In this paper, we propose a model based on rich spherical impostors, able to combine precomputed as well as dynamic procedural data, and offering seamless transitions from close instanced meshes to distant points. Our approach is architectured around an on-the-fly discrimination mechanism and intensively exploits the rough spherical geometry of the impostor proxy. In particular, we propose a new sampling mechanism to reconstruct novel views from the precomputed ones, together with a new conservative occlusion culling method, coupled with a two-pass rendering pipeline leveraging early-Z rejection. As a result, our system scales well and is even able to render sand, while supporting completely dynamic stackings.     

Dynamic Reconfiguration of Event Driven Architecture Using Reflection and Higher-Order Functions

Component-based modelling is used as the basis of a number of approaches including Enterprise Architecture and System Architecture Design. Service Oriented Architecture (SOA) is a popular component-based approach but it has been criticised as not being suficiently flexible. A more flexible alternative is Event Driven Architecture (EDA) that can support Complex Event Processing. Dynamic reconfiguration of component behaviour is attractive because it allows an architecture to be extended and modified in situ without being taken off-line, updated and redeployed. This article shows how higher-order functions and reflection can support dynamic reconfiguration and how this approach is integrated with EDA. The approach is defined as patterns expressed in a component modelling language called LEAP and validated through a case study.     

基于混合整数线性规划无人机实时航迹规划

为了解决无人机实时航迹规划问题,特别足带动力学约束条件的实时航迹规划问题,给出了基于混合整数线性规划技术在模型预测控制框架下进行无人机实时航迹规划的方法.通过将威胁区、速度、加速度以及威胁规避等约束条件转化为能够直接应用在MILP中的形式,并结合模型预测控制方法来进行规划以满足实时性要求.在威胁区的规避上,使用了二进制变量进行逻辑判断,同时,利用松弛变量的方法将威胁规避条件转变为线性形式;在速度、加速度约束条件上,使用单位圆将其约束在圆内以满足速度约束的限制.最后根据仿真计算的验证和分析,得出基于混合整数线性规划的无人机实时航迹规划的有效性.     

网络实时管理在Applet客户端的实现

本文简要介绍了目前网管系统中采用的两种客户端技术：HTML客户端和基于Swing的Applet客户端；对比了两者的优势与不足，在技术上证明了Applet客户端是实现网络实时管理的有效工具，并给出了实时性能分析曲线和告警盘的具体解决方案。     

利用RTDX技术实现计算机实时话音分析

介绍了一种利用RTDX技术作为PC同DSP之间的数据交换手段,实现计算机实时话音分析的方法及其实现,提出了PC／Windows＋RTDX＋DSP的软硬件架构模型,并将其同其它模型作了比较,着重指出了该模型的通用性和实用性。