基于稀疏字典的图像去噪方法研究

传统的去噪方法在去除噪声时,不仅会造成较大的模糊,还会影响图像的边缘信息。文章利用稀疏表示理论去噪,将包含噪声的图像分解成低频信号和高频信号,由于噪声大部分都是高频信号,于是我们利用K-SVD算法仅对高频信号进行去噪,然后将去噪后的高频信号和原有的低频信号相融合。这样能有效地降低噪声对图像影响,为下一步的图像处理打好了基础。文章通过实验对比了传统空间域和频率域的去噪方法和NL-means去噪算法。试验结果表明,文章内所述方法在去除噪声的同时,能较好的保留图像的边缘和细节信息。     

基于L1-L2混合噪声模型的图像超分辨率重建方法

文章提出了一种基于L1-L2混合噪声模型(HNM)方法用于图像、视频的超分辨重建,该方法有效利用L1范数在突变区域较好保持图像边缘信息的特点,同时兼顾了L2范数对图像平坦区域的噪声抑制特性。针对求解过程中噪声分布的改变,L1、L2范数的选择问题,提出一种自适应隶属度(AMD)方法,该方法不仅能达到理想的实验效果,而且大大地减少了迭代次数与运算时间。     

基于变分的图像恢复算法

文章旨在有效去除噪声的同时恢复图像细节和保护图像边缘,根据变分模型中φ（│ μ│）函数的选取不同,效果不同,提出了新的模型。理论可证明模型具有稳态解、算法具有收敛性。使用数值方法进行计算,用迭代算法（共轭梯度法）求解。计算结果表明,该方法可以有效恢复图像,在客观标准评价和主观视觉效果方面都有明显的改善。     

一种基于小波包分解图像的去噪方法

用小波包分解图像，对图像进行去噪，在图像去噪中能得到很好的效果。这里首先利用高斯一拉普拉斯边缘检测方法检测图像的边缘，再用小波包变换对图像的平滑区进行阈值去噪，并且对图像邻域进行平滑处理，最后把边缘图像嵌入到平滑图像。这种图像处理方法不仅可以去除图像的噪声，图像的去噪效果和清晰度也大大地得到了提高，而且图像的边缘信息也得到了很好的保持。     

基于Neyman-Pearson准则和小波模极大值的边缘检测

图像边缘是对图像进行进一步处理和识别的基础,边缘都是图像上灰度不连续的点或灰度变化剧烈的地方,所以说图像的边缘就是信号的高频部分,边缘检测就是检测信号的高频部分,是图像基本特征之一。但在实际的图像中,由于噪声的存在,边缘检测成为一个难题,所以本文提出了一种基于Neyman-Pearson准则和小波模极大值的边缘检测方法,该方法能很好地检测出图像的边缘。     

自适应双边滤波超声斑点降噪

斑点噪声压缩和组织结构保留对超声图像的临床诊断非常重要,因此文章提出一种自适应方差双边滤波超声斑点降噪方法。该方法用4方向梯度计算局部方差的变化量,以此变化量将超声图像归类为斑点噪声和组织结构,并用双边滤波滤除噪声。物理体膜图像和人类肝脏图像的实验结果表明该方法可以有效滤除斑点噪声和保持图像边缘。     

基于合成孔径雷达图像的边缘检测研究

合成孔径雷达（SAR）图像边缘信息对于图像理解有着重要意义,它既是图像分割的有效途径,又是特征提取的有效手段,也是SAR图像处理中研究的重点。文章深入分析了SAR图像的Ratios边缘检测方法,并针对Ratios算法存在边缘定位精度低的缺点,提出了一种改进的Ratios边缘改进提取算法。结果表明,方法具有较高检测正确率及较好的边缘定位精度。     

一种基于小波尺度相关性的自适应阈值去噪算法

针对图像去噪时造成的边缘模糊和细节损失问题,结合小波尺度相关性理论和经典的阈值去噪方法,提出了一种新的图像去噪算法,该算法利用不同尺度上小波系数间的相关性,对原图像和噪声对应的小波系数进行最大程度的区分,对前者予以保留,对后者则结合闽值去噪方法和该小波系数是原图像的置信程度进行处理。实验结果表明,该算法在达到很好的图像去噪的同时,能很好地保留图像的边缘和细节信息。     

一种基于距离的图像加权中值滤波方法研究

针对高压巡检车视觉获取路径信息的问题。由于噪声掩盖了电线在图像中的成像特征,给获取路径信息带来了很大的困难且影响了循迹的精度,所以提出了一种基于灰度距离加权的自适应中值滤波方法来滤除图像噪声。该方法经预先滤波后,再通过噪声检测确定噪声点及其个数,自适应的调整稀疏分布模板,最后根据模板内灰度距离而赋予各像素点不同的权重值而滤除噪声。实验结果表明,该算法所加权重值简单易实现,在很好的滤除噪声且较好的保护图像细节,还能较好的滤除图像随机噪声,使得其可很好的应用于路径图像的噪声滤除。     

基于系数优化的图像边缘增强误差扩散方法

传统误差扩散法所生成的半色调图像容易产生"蠕虫"效应及边缘模糊的现象。因此,本文提出了一种基于系数优化的图像边缘增强误差扩散方法,该方法以Shiau-Fan所提出的新加权误差传递算法为基础,采用canny算子来检测图像的边缘且沿边缘方向不断优化其误差滤波器的系数,从而达到增强半色调图像边缘及弱化"蠕虫"现象的目的。     

林业可视化技术推广模式及其创新

一种技术只有应用于实际才算是真正拥有价值。林业可视化是一门鲜为人知新兴技术,虽然在某些领域有了一些应用,但其规模尚小,深度尚浅。因此,该技术的推广应用还有很大的发展空间。本文从林业可视化技术推广模式的角度出发,对林业可视化技术的推广进行了深入研究,借助其他科学技术的推广模式,结合林业可视化技术的自身特点,对林业可视化技术的推广模式进行了总结分类,并且提出了创新。     

集群企业创业传导仿真的构建分析

集群企业二次创业已成为产业集群转型升级的重要路径,而集群企业网络影响了集群企业的二次创业行为。本文从复杂网络视角出发,在对集群企业网络与复杂网络模型之间耦合关系的探讨基础上,分析了用仿真方法构建集群企业网络,探讨了如何模拟集群企业网络中创业行为的传导,为刻画现实的集群企业创业行为传导过程和效果提供了新的分析思路。     

三重扩散工艺的改进

扩散技术在工业制作的各个方面有重大的作用,在电子技术上的应用以晶体管最为显著。本文就简要介绍如何应用三重扩散工艺生产大功率晶体管,并对目前生产工艺中的一些不足之处进行改进,以更好的提高大功率晶体管的质量。     

An empirical study of e-government diffusion in Fiji: a holistic and integrative approach

Pacific Island countries (PICs) are facing certain unique challenges in e-government diffusion due to their geographic, cultural, social, and economic environments. However, very few rigorous empirical studies have been published about e-government diffusion issues in the region. To address the knowledge gap, we conducted a case study of Fiji. Drawing on Fountain’s technology enactment framework as a theoretical lens, we identified and investigated in a holistic and integrative way the major issues and challenges which influenced e-government diffusion and outcomes in Fiji. The study has significant implications both for research and practices not only in Fiji but also in the PICs.     

基于局部网络效应的创新扩散分析

在考虑局部网络效应的基础上,采用微观扩散模型仿真分析创新扩散。研究表明,局部网络效应使得创新潜在采纳个体决策相互影响,从而导致这种采纳者之间微观交互模式所呈现的宏观社会网络结构影响创新扩散的速度和程度。即使某些创新收益非常高,也有可能因为创新潜在采纳个体之间交互模式的影响而导致其扩散最终失败。同时早期创新采纳个体一般具有社会联系比较广泛,自身采纳阈值较高,自身局部网络效应强度较小,更关注创新内在价值等创新特征。     

Mandate Versus Championship: Vertical government intervention and diffusion of innovation in public services in authoritarian China

This research focuses on innovation and its diffusion in public services in authoritarian China. A mechanism between vertical government intervention and diffusion of innovation in public services is established by conducting a comparative case study between Sichuan and Tianjin. Administrative commands facilitate the formation of the ‘mandatory policy diffusion’ that rapidly diffuses policy instruments. Competition in the performance evaluation-based personnel system contributes to the formation of ‘championship policy diffusion’, which leads to the divergence of policy instruments in neighbouring local governments. Therefore, classic theoretical hypotheses on geographical proximity, competition, and vertical intervention concerning innovation diffusion need to be modified.     

Why do statistics journals have low impact factors?

In this paper, we answer the question why statistics journals get lower impact factors than journals in other disciplines. We analyze diffusion patterns of papers in several journals in various academic fields. To obtain insights into the diffusion of the citation counts of the papers, the data are analysed with the Bass model, leading to values for the time-to-peak that can be used to compare the speeds of diffusion paper citations of the different disciplines. Estimation results show that for statistics journals, it takes significantly more years to reach their peak. To further investigate diffusion, we also compute the percentages of the total number of citations a paper has after 2 or 3 years. Again, it appears that statistics journals have slower citation diffusion than journals in other disciplines. We conclude with some suggestions to reduce this disparity.     

Good neighbours matter: economic geography and the diffusion of human rights

Using multi-country panel data, this paper investigates the geopolitical and economic aspects of human rights performance. Human rights performance depends on the relative levels of economic development and spatial proximity to ‘good’ and ‘bad’ neighbours. The paper tests for basic effects of income, and applies spatial weighting models, to analyse the neighbours’ impact on human rights levels, treating this impact as partly endogenous. It takes into account size and distance when comparing each country’s human rights performance with what would be predicted from a weighted average of its neighbours’ performance. There are (1) geographical clusters and (2) size and proximity effects for human rights performance.     

Are Scottish firms meeting the ICT challenge? Results from a National Survey of Enterprise

This paper examines the diffusion of Information Communication Technology (ICT) into firms in Scotland and northern England. Data concerning the adoption of micro-computers, e-mail, the Internet, ISDN, in-house website, and the automation of business functions is analysed by industry, firm size, firm age, rate of growth, export involvement, and innovation activity. The results from a sample of 1347 firms found an overall increase in the adoption of ICT in firms between 1998 and 2001. The increase is led by older and larger firms in comparison to younger and smaller firms. In addition, the diffusion of the automation of business functions was found to be sequential, from generic to specific applications. Further increases in the diffusion of ICT are likely to come from upgrading existing equipment and increases in the range of business functions that can be automated. Finally, although use of the Internet and web-based trading has increased in the sample firms, this has not replaced traditional marketing and sales. The results support the view that more firms are taking up the challenge of using ICT in-house, with older and larger firms leading the way ahead of younger and smaller firms.