Quantization error and dithering

Digital storage and transmission promise noise-free images, but it is important to keep in mind that even digital is not perfect. Digital images have their own sources of noise: round-off error and quantization error. Whenever you do any sort of image arithmetic, such as contrast enhancement or compositing, you get roundoff error. In fact, since the arithmetic is often done in only X-bit accuracy, sometimes the round-off error can be substantial. You get quantization error, on the other hand, whenever you go from an analog signal to a digital signal or whenever you go from a high color-resolution signal (for example, 24 bits per pixel) to a low resolution signal (for example, 8 bits per pixel). The author considers the quantization error from analog to digital     

Quantization error and design of digital control systems

The evolution of the error caused by the amplitude quantization in digital control systems can be studied by introducing bounded disturbances at the points of the quantization. Such a formulation provides an upper bound for the actual quantization error. This paper introduces the problem of the quantization error in the light of the discrete-time optimum control theory. The determination of the worst possible effect of the error due to quantization is reduced to solving a common optimum control problem; the "worst effect" is defined as the maximum of a chosen performance criterion. For a design, a method is proposed for the minimization of the worst effect of the quantization error. Computational solutions illustrate the presentation.     

基于量化误差与分形理论的高计算效率无监督聚类研究*

已有的矢量聚类算法需学习较多的复杂数据方可获得较好的聚类效果,而对于多维的大数据性能较弱,对此,提出一种基于量化误差与分形理论的高计算效率无监督聚类算法。首先,为数据集建立量化误差的参数化模型,基于数据集的空间结构获得数据集的率失真曲线;然后,通过对率失真曲线的估算,获得数据空间的有效维度;最终,利用分形理论,通过搜索数据集的量化模型参数获得目标数据集的最优类簇数量。实验结果表明,本文的量化误差参数化模型可较好地估算数据集的有效维度,同时,本算法对数值型数据集的最优类簇估算与计算效率优于已有的矢量聚类算法。     

数字闭环石英挠性加速度计量化误差分析与实验研究

数字闭环方案可以减小模拟伺服方案中A／D转换环节所引入的精度损失,而量化误差是影响数字闭环石英挠性加速度计系统精度的重要因素之一。介绍了数字闭环石英挠性加速度计检测系统的工作原理,讨论了AD,DA位数的选择依据,基于现有硬件检测电路,软件上采取AD,DA降位使用的方法分析量化误差对闭环系统精度的影响。搭建数字闭环实验系统,实验结果表明：在一定范围内,随着AD,DA位数的降低,量化误差影响增大,闭环系统精度降低,与理论分析相吻合,为数字闭环加速度计的精度改善提供必要的理论指导和实验基础。     

Characterization of minimum error linear coding with sensory and neural noise

Robust coding has been proposed as a solution to the problem of minimizing decoding error in the presence of neural noise. Many real-world problems, however, have degradation in the input signal, not just in neural representations. This generalized problem is more relevant to biological sensory coding where internal noise arises from limited neural precision and external noise from distortion of sensory signal such as blurring and phototransduction noise. In this note, we show that the optimal linear encoder for this problem can be decomposed exactly into two serial processes that can be optimized separately. One is Wiener filtering, which optimally compensates for input degradation. The other is robust coding, which best uses the available representational capacity for signal transmission with a noisy population of linear neurons. We also present spectral analysis of the decomposition that characterizes how the reconstruction error is minimized under different input signal spectra, types and amounts of degradation, degrees of neural precision, and neural population sizes.     

平均误差向量加速的[K]-Means色彩量化方法

针对[K]-Means色彩量化方法在运行时间上过于冗长的问题,提出一种用平均误差向量加速的色彩量化方法。随机生成[K]种色彩作为初始的调色盘,用该调色盘对欲量化的图像进行一次量化。根据量化后的版本,计算其每个颜色分量的量化误差,获得平均误差向量。用该平均误差向量对调色盘进行更新,获得另一更优的调色盘。通过若干次迭代运算,获得最终收敛的调色盘,并用该调色盘进行最后的色彩量化。实验结果表明,该加速算法能对[K]-Means量化方法平均加速70~150倍,同时,原有[K]-Means方法的量化效果还得到了保持。     

Dithered Color Quantization

Image quantization and digital halftoning are fundamental problems in computer graphics, which arise when displaying high-color images on non-truecolor devices. Both steps are generally performed sequentially and, in most cases, independent of each other. Color quantization with a pixel-wise defined distortion measure and the dithering process with its local neighborhood optimize different quality criteria or, frequently, follow a heuristic without reference to any quality measure.
In this paper we propose a new method to simultaneously quantize and dither color images. The method is based on a rigorous cost-function approach which optimizes a quality criterion derived from a generic model of human perception. A highly efficient algorithm for optimization based on a multiscale method is developed for the dithered color quantization cost function. The quality criterion and the optimization algorithms are evaluated on a representative set of artificial and real-world images as well as on a collection of icons. A significant image quality improvement is observed compared to standard color reduction approaches.     

基于高斯格型矢量量化的导谱频率参数量化方法

为了降低宽带导抗谱频率(ISF)参数矢量量化器的运算复杂度和存储空间,提出了一种随机矢量-高斯格型矢量两级矢量量化模型,研究了量化模型的联合量化算法和码本训练算法.基于上述量化模型设计了一种宽带ISF参数预测式两级矢量量化器,量化算法的第一级使用随机矢量量化第二级采用高斯格型矢量量化.实验结果表明,ISF参数预测式两级矢量量化器在45比特/帧可以达到透明量化,相比于ITU-T G.722.2中的分裂-多级矢量量化(S-MSVQ)方法,该方法降低了41.7%的存储空间和3..1%的运算复杂度.     

JPEG2000中的自适应量化研究

根据人类视觉特性和小波子带图像分布特性及子带系数对保存边缘、纹理信息的重要性程度,提出一种基于图像纹理分类的自适应量化编码算法,在同一子带内根据小波系数对保存边缘、纹理信息的重要性程度,将小波高频子带内划出平坦块、纹理块和边缘块,对其采用不同的量化间隔,并以JPEG2000为平台进行了实验.结果证明,采用该算法,可以在低码率下较好地保存图像的边缘和纹理特性的信息,重构图像的主观质量有一定的提高.     

Confirming configurations in EFSM testing

We investigate the problem of configuration verification for the extended FSM (EFSM) model. This is an extension of the FSM state identification problem. Specifically, given a configuration ("state vector") and an arbitrary set of configurations, determine an input sequence such that the EFSM in the given configuration produces an output sequence different from that of the configurations in the given set or at least in a maximal proper subset. Such a sequence can be used in a test case to confirm the destination configuration of a particular EFSM transition. We demonstrate that this problem could be reduced to the EFSM traversal problem, so that the existing methods and tools developed in the context of model checking become applicable. We introduce notions of EFSM projections and products and, based on these notions, we develop a theoretical framework for determining configuration-confirming sequences. The proposed approach is illustrated on a realistic example.     

Quantization Error in Stereo Imaging systems

In this paper a stochastic analysis of the quantization error in a stereo imaging system has been presented. Further the probability density function of the range estimation error and the expected value of the range error magnitude are derived in terms of various design parameters. Further the relative range error is proposed.     

Quantization errors in digital control systems

An analytic study was conducted to develop a quantitative expression for the errors due to quantization (round-off) in digital control systems. An upper bound on the error in the output that was caused by quantization became available for two different sampling rates by describing a typical digital control system in terms of its state variables. The amount of the error introduced in the output by any one quantization operation was found to depend upon the sampling rate, system time constants, and the form of the controller chosen for system compensation. For a third-order plant, preceded by a zero-order hold circuit and compensated by a controller that provided a response dominated by a closed-loop pole pair with a damping ratio of 0.7, the error due to quantization was determined to be approximately ten times the value of the quantizing level. The error was reduced to six times the quantizing level by increasing the sampling rate by a factor of 2, to 20 per second. The results of the study furnished a convenient procedure for analyzing quantization in a digital control system and led to a specification on the precision required in the digital controller and the necessary encoding equipment.     

基于自组织特征映射的图象矢量量化

本文就基于自组织特征映射的图象矢量量化编码做了初步的探讨,得出一些结论。在矢量量化中,码本性能的好坏对重建的图像有直接的影响。我们利用自组织特征映射(SOFM)网络进行聚类,实现了图像矢量码本的生成,然后再根据矢量量化(VQ)编码原理将图像重建。该方法可以达到较高的压缩比,实现了图像压缩。并且,就不同条件下的图像作了对比。     

Object configurations in software engineering databases

We consider the logical organization of database support for software engineering applications, which has to cope with such requirements as the adequate support of object composition and versioning, the derivation of (consistent) configurations, and the provision of application-specific transactions. A new approach to the organization, manipulation and management of multiversion objects in CASE databases is described and investigated, which clearly distinguishes four relevant concepts: documents, versions, configurations and databases. The approach is formally made precise using the concept ofAND/OR graphs, which renders it possible to cast operations on any of the concepts in terms of graph operations. The levels of abstraction distinguished and maintained throughout the exposition give rise to an appropriate transaction concept: transactions involve graph operations on specific types of objects only; conflicts between transactions can thus be easily identified, and consistency of objects is easily maintained.     

MatLab在图像矢量量化中的应用

从矢量量化的特点和码本设计方法出发,探讨MatLab在矢量量化中的应用,利用MatLab编程实现图像矢量量化,并给出了程序运行结果,通过对运行结果进一步分析,效果良好,为图像矢量量化提供了一个新的途径.     

Testing component compatibility in evolving configurations

Software components are increasingly assembled from other components. Each component may further depend on others, and each may have multiple active versions. The total number of configurations—combinations of components and their versions—in use can be very large. Moreover, components are constantly being enhanced and new versions are being released. Component developers, therefore, spend considerable time and effort doing compatibility testing—determining whether their components can be built correctly for all deployed configurations–both for existing active component versions and new releases. In previous work we developed Rachet, a distributed, cache-aware mechanism to support large-scale compatibility testing of component-based software with a fixed set of component versions.In this paper, we observe that it is too expensive to perform compatibility testing from scratch each time a new version of a component is released. We thus add a new dimension to Rachet: to perform incremental and prioritized compatibility testing. We describe algorithms to compute differences in component compatibilities between current and previous component builds, a formal test adequacy criterion based on covering the differences, and cache-aware configuration sampling and testing methods that attempt to reuse effort from previous testing sessions. Because testers are often interested in focusing test effort on newly released and modified components and their versions, we have developed a prioritization mechanism that enhances compatibility testing by examining the configurations that test new or modified component versions first, while also distributing the work over a cluster of machines. We evaluate our approach using the 5-year evolution history of a scientific middleware component. Our results show that our methods can increase performance significantly over Rachet’s previous retest-all approach and also tests important component compatibilities early in the overall testing process, making the process of compatibility testing practical for evolving components.     

CoQuWeiP--使用加权积的色彩量化算法

提出一种新颖的色彩量化算法——CoQuWeiP．该算法通过设立像素个数和色彩距离的权重,综合考虑了色彩层次感和关键细节的要求,并可以调节权重以满足不同色彩量化任务对色彩层次感和关键细节的不同要求．实验表明,文中算法在调节量化结果方面是有效的,量化性能较好．     

程序局部性的量化分析

给出与平台无关的局部性量化方法,从空间局部性和时间局部性2个角度,量化SPEC2000测试基准程序,以及这些程序的数据段、代码段和堆栈段。时间和空间局部性组成的二维局部性分布直观地展示了基准测试程序的局部性。实验结果表明,程序数据局部性主要由堆段的局部性决定,堆段的局部性最差,栈的局部性最优。