Feature tracking and matching in video using programmable graphics hardware

This paper describes novel implementations of the KLT feature tracking and SIFT feature extraction algorithms that run on the graphics processing unit (GPU) and is suitable for video analysis in real-time vision systems. While significant acceleration over standard CPU implementations is obtained by exploiting parallelism provided by modern programmable graphics hardware, the CPU is freed up to run other computations in parallel. Our GPU-based KLT implementation tracks about a thousand features in real-time at 30 Hz on 1,024 × 768 resolution video which is a 20 times improvement over the CPU. The GPU-based SIFT implementation extracts about 800 features from 640 × 480 video at 10 Hz which is approximately 10 times faster than an optimized CPU implementation.     

Performance evaluation of a 3D multi-view-based particle filter for visual object tracking using GPUs and multicore CPUs

This paper presents a deep and extensive performance analysis of the particle filter (PF) algorithm for a very compute intensive 3D multi-view visual tracking problem. We compare different implementations and parameter settings of the PF algorithm in a CPU platform taking advantage of the multithreading capabilities of the modern processors and a graphics processing unit (GPU) platform using NVIDIA CUDA computing environment as developing framework. We extend our experimental study to each individual stage of the PF algorithm, and evaluate the quality versus performance trade-off among different ways to design these stages. We have observed that the GPU platform performs better than the multithreaded CPU platform when handling a large number of particles, but we also demonstrate that hybrid CPU/GPU implementations can run almost as fast as only GPU solutions.     

Fermi架构下的时域高斯滤波并行算法

为提高图形图像处理中高斯滤波算法模块的计算速度,将高斯滤波与Fermi平台相结合,设计了一种高斯滤波时域的并行算法。数据测试结果显示,与基于CPU的实现相比,采用Fermi架构的GPU处理不仅可以得到误差精度小于0.0001的计算结果,而且可以取得较大的加速效果。在数据规模为512×112×128和滤波窗口大小为11的情况下能够达到约210倍的加速效果。     

Denoising methods for thermomechanical decomposition for quasi-equilibrium molecular dynamics simulations

In this work, we applied robust denoising methods well established in the signal processing field for the thermomechanical decomposition of velocity data obtained from molecular dynamics (MD) simulations. In the decomposition, the atomic velocity was assumed to be the sum of the mechanical velocity and the thermal velocity, which can be linked to the stress and temperature field at the continuum scale, respectively. For the quasi-equilibrium process, with the thermal velocity treated as the Gaussian distributed stationary noise with zero mean and a positive variance that is linearly proportional to temperature and mechanical velocity as the clean signal, the velocity decomposition can be recasted into a denoising problem, for which powerful denoising methods have been developed to estimate a clean signal from noisy data. We investigated the widely-used linear parametric real-domain, linear nonparametric Fourier-based, and nonlinear nonparametric wavelet-based denoising methods, first on their theoretical properties and then made comparsion among them for denosing some synthetic noisy 1-dimensional (1D) data generated from MD simulations. The nonlinear wavelet-based thresholding estimator possessed better optimality properties than the other estimators, and also outperformed the other estimators in the synthetic data test. A further test comparing the various denoising methods for an adiabatic shear crack nucleation and propagation process simulated using MD simulations showed better performance by the wavelet-based denoisng method. Results from this work reveal good potential of applying wavelet-based denoising method to the study of thermomechanical processes simulated using MD simulations.     

GPU-based anisotropic diffusion algorithm for video image denoising

Image filtering is the process of removing noise which perturbs image analysis methods. In some applications like segmentation, denoising is intended to smooth homogeneous areas while preserving the contours. Real-time denoising is required in a lot of applications like image-guided surgical interventions, video analysis and visual serving. This paper presents an anisotropic diffusion method named the Oriented Speckle Reducing Anisotropic Diffusion (OSRAD) filter. The OSRAD works very well for denoising images with speckle noise. However, this filter has a powerful computational complexity and is not suitable for real time implementation. The purpose of this study is to decrease the processing time implementation of the OSRAD filter using a parallel processor through the optimization of the graphics processor unit. The results show that the suggested method is very effective for real-time video processing. This implementation yields a denoising video rate of 25 frames per second for 128 × 128 pixels. The proposed model magnifies the acceleration of the image filtering to 30 × compared to the standard implementation of central processing units (CPU). A quantitative comparison measure is given by parameters like the mean structural similarity index, the peak signal-to-noise ratio and the figure of merit. The modified filter is faster than the conventional OSRAD and keeps a high image quality compared to the bilateral filter and the wavelet transformation.     

侧扫声呐图像的3维块匹配降斑方法

斑点噪声是影响侧扫声呐图像质量的主要因素,降斑处理对侧扫声呐图像的判别与分析非常重要。针对侧扫声呐图像自身特性和斑点噪声分布特点,提出一种基于3维块匹配(BM3D)的降斑方法。根据海底散射模型,得到侧扫声呐图像斑点噪声的瑞利分布模型,然后通过高斯光滑函数幂变换将瑞利分布的噪声转化为高斯分布,通过对数变换将乘性噪声转变为加性噪声,再进行自适应的BM3D滤波,最后采用逆变换得到降斑图像。实验结果表明,该方法在降噪、边缘和纹理保持等方面均优于空间域、小波域、Curvelet域的一些降斑方法。     

A GPU implementation for LBG and SOM training

Vector quantization (VQ) is an effective technique applicable in a wide range of areas, such as image compression and pattern recognition. The most time-consuming procedure of VQ is codebook training, and two of the frequently used training algorithms are LBG and self-organizing map (SOM). Nowadays, desktop computers are usually equipped with programmable graphics processing units (GPUs), whose parallel data-processing ability is ideal for codebook training acceleration. Although there are some GPU algorithms for LBG training, their implementations suffer from a large amount of data transfer between CPU and GPU and a large number of rendering passes within a training iteration. This paper presents a novel GPU-based training implementation for LBG and SOM training. More specifically, we utilize the random write ability of vertex shader to reduce the overheads mentioned above. Our experimental results show that our approach can run four times faster than the previous approach.     

能量回归尺度空间滤波法在图像去噪中的应用

如何有效地增强图像边缘信息,消除图像的噪声是图像处理和计算机视觉研究中的基本问题。利用多尺度二进小波变换,对2维噪声图像进行了研究,提出了能量回归尺度空间滤波方法。利用一些图像进行了数值实验,实验结果表明,能量回归尺度空间滤波法可以较好地保留图像边缘特征,较多地去除噪声。与典型的小波萎缩去噪方法：“硬阈值”滤波法、“软阈值”滤波法相比较,能量回归滤波算法的峰值信噪比（PSNR）提高了2~3dB,从而证实了能量回归尺度空间滤波方法具有良好的去噪性能。     

具有自适应窗口的双变量模型图像去噪方法

针对小波域去噪方法BiShrink-local(双变量萎缩局部方差估计)会造成图像的细节信息丢失的问题,给出一种具有自适应窗口的双变量模型图像去噪方法.该方法一方面继承原方法的优点,另一方面又利用区域生长原理,通过判断图像的小波系数值是否属于同质,从而更好地区分噪声和细节信息.通过实验表明,该方法能对含有高斯噪声的图像进行较好地去噪,同时在保持细节方面优于原来的方法.     

CUDA optimization strategies for compute- and memory-bound neuroimaging algorithms

As neuroimaging algorithms and technology continue to grow faster than CPU performance in complexity and image resolution, data-parallel computing methods will be increasingly important. The high performance, data-parallel architecture of modern graphical processing units (GPUs) can reduce computational times by orders of magnitude. However, its massively threaded architecture introduces challenges when GPU resources are exceeded. This paper presents optimization strategies for compute- and memory-bound algorithms for the CUDA architecture. For compute-bound algorithms, the registers are reduced through variable reuse via shared memory and the data throughput is increased through heavier thread workloads and maximizing the thread configuration for a single thread block per multiprocessor. For memory-bound algorithms, fitting the data into the fast but limited GPU resources is achieved through reorganizing the data into self-contained structures and employing a multi-pass approach. Memory latencies are reduced by selecting memory resources whose cache performance are optimized for the algorithm's access patterns. We demonstrate the strategies on two computationally expensive algorithms and achieve optimized GPU implementations that perform up to 6× faster than unoptimized ones. Compared to CPU implementations, we achieve peak GPU speedups of 129× for the 3D unbiased nonlinear image registration technique and 93× for the non-local means surface denoising algorithm.     

局部均值噪声估计的盲3维滤波降噪算法

目的 图像在获取和传输的过程中很容易受到噪声的干扰,图像降噪作为众多图像处理系统的预处理模块在过去数十年中得到了广泛的研究。在已提出的降噪算法中,往往采用加性高斯白噪声模型AWGN（additive white Gaussian noise）为噪声建模,噪声水平（严重程度）由方差参数控制。经典的BM3D 3维滤波算法属于非盲降噪（non-blind denoising algorithm）算法,在实际使用中需要由人工评估图像噪声水平并设置参数,存在着噪声评估值随机性大而导致无法获得最佳降噪效果的问题。为此,提出了一种新的局部均值噪声估计（LME）算法并作为BM3D算法的前置预处理模块。方法 本文专注于利用基于自然统计规律（NSS）的图像质量感知特征和局部均值估计技术构建图像噪声水平预测器,并通过它高效地获得噪声图像中准确的噪声水平值。关于自然场景统计方面的研究表明,无失真的自然场景图像在空域或者频率域上具有显著的统计规律,一旦受到噪声干扰会产生规律性的偏移,可以提取这些特征值作为反映图像质量好坏的图像质量感知特征。另外,局部均值估计因其简单而高效率的预测特性被采用。具体实现上,在具有广泛代表性且未受噪声干扰图像集合上添加不同噪声水平的高斯噪声构建失真图像集合,然后利用小波变换对这些失真图像进行不同尺度和不同方向的分解,再用广义高斯分布模型（GGD）提取子带滤波系数的统计信息构成描述图像失真程度的特征矢量,最后用每幅失真图像上所提取的特征矢量及对其所施加的高斯噪声水平值构成了失真特征矢量库。在降噪阶段,用相同的特征提取方法提取待降噪的图像的特征矢量并在失真特征矢量库中检索出与之类似的若干特征矢量及它们所对应的噪声水平值,然后用局部均值法估计出待降噪图像中高斯噪声大小作为经典BM3D算法的输入参数。结果 改进后的BM3D算法转换为盲降噪算法,称为BM3D-LME（block-matching and 3D filtering based on local means estimation）算法。准确的噪声估计对于诸如图像降噪,图像超分辨率和图像分割等图像处理任务非常重要。已经验证了所提出噪声水平估计算法的准确性、鲁棒性和有效性。结论 相对人工进行噪声估计,LME算法能够准确、快速地估算出任意待降噪图像中的噪声大小。配合BM3D算法使用后,有效提高了它的实际降噪效果并扩大它的应用范围。     

改进的曲率滤波强噪声图像去噪方法

目的 医学影像获取和视频监控过程中会出现一些恶劣环境,导致图像有许多强噪声斑点,质量较差。在处理强噪声图像时,传统的基于变分模型的算法,因需要计算高阶偏微分方程,计算复杂且收敛较慢;而隐式使用图像曲率信息的曲率滤波模型,在处理强噪声图像时,又存在去噪不完全的缺陷。为了克服这些缺陷,在保持图像边缘和细节特征的同时去除图像的强噪声,实现快速去噪,提出了一种改进的曲率滤波算法。方法 本文算法在隐式计算曲率时,通过半窗三角切平面和最小三角切平面的组合,用投影算子代替传统曲率滤波的最小三角切平面投影算子,并根据强噪声图像存在强噪声斑点的特征,修正正则能量函数,增添局部方差的正则能量,使得正则项的约束更加合理,提高了算法的去噪性能,从而达到增强去噪能力和保护图像边缘与细节的目的。结果 针对多种不同强度的混合噪声图像对本文算法性能进行测试,并与传统的基于变分法的去噪算法（ROF）和曲率滤波去噪等算法进行去噪效果对比,同时使用峰值信噪比（PSNR）和结构相似性（SSIM）作为滤波算法性能的客观评价指标。本文算法在对强噪声图像去噪处理时,能够有效地保持图像的边缘和细节特征,具备较好的PSNR和SSIM,在PSNR上比ROF模型和曲率滤波算法分别平均提高1.67 dB和2.93 dB,SSIM分别平均提高0.29和0.26。由于采用了隐式计算图像曲率,算法的处理速度与曲率滤波算法相近。结论 根据强噪声图像噪声特征对曲率滤波算法进行优化,改进投影算子和能量函数正则项,使得曲率滤波算法能够更好地适用于强噪声图像,实验结果表明,该方法与传统的变分法相比,对强噪声图像去噪效果显著。     

Object oriented framework for real-time image processing on GPU

General purpose computation on graphics processing unit (GPGPU) provides a significant gain in terms of the processing time compared with CPU. Images are particularly good subjects for massive parallel implementations on GPU. Thus, the processing time can be improved for computer vision and image/video processing algorithms. However, GPGPU has a fairly complex integration process in a framework and they evolve very rapidly. In this paper, we present a framework that provides all the desired primitives related to GPGPU-based image processing algorithms, which makes it easy and straightforward for the user to exploit. The proposed framework is object-oriented, and it utilizes design patterns. The user can benefit from all the advantages of object-oriented programming, such as code reusability/extensibility, flexibility, information hiding, and complexity hiding. This makes it possible to rapidly integrate new technologies and functionality as they appear.     

基于BM3D的复杂纹理区域图像去噪

纵观图像去噪领域,目前也出现了一些效果较好,并且应用广泛的图像去噪算法,例如频域滤波中的小波分解、遗传算法(GA)、字典学习算法(K-SVD)、非局部滤波(Non-Local Means,NLM)等等,可以说在单图像去噪上效果最被认可的还是块匹配3D去噪(BM3D)。但在该方法应用的过程中,还是出现了一些需要改进的实际问题。特别是在图像复杂纹理区域的去噪上,由于BM3D去噪的结构特性,导致了在该区域上并没有得到较好的效果。所以在该文提出了一种结合TV模型和改进的Prewitt边缘检测的混合去噪方法。最后的实验结果也表明了改进的BM3D方法相对于原始的方法还是取得了不错的提升。     

基于图结构滤波的图像去噪

在图像处理领域, 图像去噪是一项极具挑战性的任务. 图信号理论的发展为我们解决这一问题提供了新的视角. 本文研究了基于图信号方法的权重矩阵与拉普拉斯矩阵, 将它们用于图像去噪的目标函数, 这两个矩阵可以很好地定义观测图像与期望图像之间的内在联系. 在提出去噪目标函数的基础上, 我们给出了最优解和一种迭代的快速求解算法. 实验表明, 该方法优于BM3D和WNNM等前沿的去噪方法.     

基于带权核范数最小化和混合高斯模型的去噪模型

非局部自相似性（NSS）先验在图像恢复中发挥重要作用,如何充分利用这一先验提高图像恢复性能仍值得深入研究,提出一种基于带权核范数最小化和混合高斯模型的去噪模型。首先,采用混合高斯模型（GMM）对无噪声的自然图像非局部自相似图像块进行训练,再用训练好的混合高斯模型指导退化的图像产生非局部自相似图像块组;然后,结合带权的核范数最小化技术实现图像的去噪,并对模型的保真项进行一般性扩展,给出收敛的求解算法。仿真实验表明,所提方法与基于3D滤波的块匹配（BM3D）算法、同时稀疏编码学习（LSSC）算法和带权的核范数最小化（WNNM）模型相比,峰值信噪比（PSNR）提高0.11~0.49 dB。     

基于离群点检测的图形图象噪声滤除算法

图形图象噪声过滤与修正,在媒体制作、图象分析与信息提取中起着十分重要的作用.虽然基于小波变换的算法能够对高斯噪声进行较好的滤噪处理,但对于随机分布于图象中的各种非高斯噪声仍没有普遍适用的滤噪方法.为了对这种随机分布于图象中的噪声进行有效的检测与滤除,采用对数字图象像素进行解析化描述的方法,从离群点检测的角度给出噪声的定义,并在此基础上构造了相应的图象噪声检测与滤除算法.实验结果表明,这一新方法对图象类型具有广泛的适应性和较好的噪声滤除效果,在大规模图形图象处理应用中具有实用价值.     

基于LS-SVM的图像去噪方法

支持向量机是一种基于统计学习理论的机器学习方法，该方法已广泛用于解决分类问题和回归问题。文中将最小二乘支持向量机应用于图像去噪中，并同小波去噪及中值滤波进行了比较分析。仿真结果表明，该方法能较好的保存图像细节，并具有很好的泛化能力。     

基于SAR统计参数的4f系统小波阈值去噪方法

针对典型4f信息光学系统的低通特性和高斯白噪声性,引入合成孔径雷达(SAR)领域中常用的统计参数标准,提出一种小波阈值去噪方案。通过实际和仿真实验验证,该方案能有效滤除噪声,同时较好地保留有用细节信息,适合类似于4f信息光学系统的含有高斯白噪声和明显低通滤波效果的应用场合。     

基于形状自适应PCA的三维块匹配图像去噪

三维块匹配法中3-D变换真实信号的稀疏表达能力较弱,针对该问题,提出一种关于图像去噪的三维块匹配(BM3D)改进算法。采用形状自适应的图像块(邻域)代替BM3D算法中的平方窗图像块,对3-D变换处理的形状自适应图像块进行PCA变换。实验结果证明,该算法能够有效去除图像的高斯噪声,提高图像的峰值性噪比和结构相似度,且在保持图像的边缘等细节信息方面性能较好,图像视觉效果有明显改善。