改进的局部最小像素先验遥感图像盲复原算法

为了解决遥感图像盲复原时模糊核估计不准确、复原图像存在振铃效应的问题,提出改进的局部最小像素先验遥感图像盲复原算法。该算法首先引入极端通道先验与局部最小像素先验结合,对图像的强度进行更好的约束,有利于得到更好的潜在清晰图像;然后采用基于梯度的方法估计模糊核,模糊核估计与中间潜在清晰图像估计交替迭代进行,获得较为理想的模糊核;最后引入联合双边滤波器,采用改进的拉普拉斯与正则化图像复原算法抑制图像复原的振铃效应。实验结果表明,本文方法对遥感图像复原效果较好,恢复的图像边缘清晰,振铃伪影得到抑制且模糊核较为理想;客观评价指标峰值信噪比(PSNR)较前沿复原算法平均提高约1.40 dB,结构相似度(SSIM)平均提高约0.02。     

离焦模糊图像的自适应滤波及逆滤波器复原

在微操作中,显微视觉系统获取的图像通常是离焦模糊图像.根据最小二乘原理和回归模型设计自适应滤波器,用于消除图像噪声,提高图像的信噪比;离焦模糊图像的退化模型可用圆盘函数描述,利用模糊图像频域的零点位置来估计圆盘函数的模糊参数;采用基于简化Wiener滤波的逆滤波器方法对模糊图像进行复原.对算法进行了仿真和实验分析,结果表明,该方法能够以较少的运算时间代价获取较好的复原效果.     

稀疏先验型的大气湍流退化图像盲复原

图像盲复原是仅从降质图像就恢复出模糊核和真实锐利图像的方法,由于其病态性,通常需要加入图像先验知识约束解的范围。针对传统的图像梯度l2和l1范数先验不能真实刻画自然图像梯度分布的特点,本文将图像梯度稀疏先验应用于单帧大气湍流退化图像盲复原中。先估计模糊核再进行非盲复原,利用分裂Bregman算法求解相应的非凸代价函数。仿真实验表明,与总变分先验(l1范数)相比,稀疏先验有利于模糊核的估计、产生锐利边缘和去除振铃等,降低了模糊核的估计误差从而提高了复原质量。最后对真实湍流退化图像进行了复原。     

基于散斑图像梯度域的数字图像相关盲去模糊方法EI北大核心CSCD

针对复杂测量环境或高动态测量过程中出现的运动模糊问题,提出了一种灰度稀疏先验与参考图像梯度域先验相结合的散斑图像盲去模糊方法。该方法以灰度直方图峰值的L 0范数与参考图像梯度域分布建立优化函数正则项,使用二次分裂方法估计清晰图像,再以交替迭代的方式进行卷积核细化。在模糊核估计完成后,使用Richardson-Lucy非盲去卷积方法完成散斑图像的复原。实验结果证明:所提出的散斑图像盲去模糊方法与针对自然图像与文本图像的经典方法相比,获得了更优的图像去模糊效果,并提高了数字图像相关测量精度与鲁棒性。     

混合梯度稀疏先验约束下的图像盲复原

图像盲复原旨在无参考的情况下准确估计模糊核并恢复潜在的清晰图像。现有研究成果表明,利用全变分模型对高阶图像梯度先验约束进行描述可以有效抑制复原图像中产生的阶梯效应。本文在实验观察和研究的基础上,提出了采用稀疏先验约束模型对盲复原过程进行正则化的方法,以获得更佳的图像复原效果。该方法利用图像高阶梯度的稀疏性,通过与低阶梯度相结合来构造混合梯度正则项。同时,在正则项中引入基于图像熵的自适应因子,来调节迭代优化过程中两类梯度先验的比例,以此获得更好的收敛性。仿真与实验证明,与现有图像盲复原先进方法相比,本文方法具有更优越的图像复原性能。     

一种基于空间自适应的盲图像复原算法

针对非负有限支持域约束递归逆滤波（NAS-RIF）图像盲复原算法对噪声敏感问题，提出一种基于空间自适应的改进算法。利用最优阈值估计图像支持域，在代价函数中引入空间自适应加权因子和尺度因子，结合正则化技术对图像复原进行局部控制，并对不同信噪比和背景的图像进行了实验。结果表明，在较低信噪比情况下，改进NAS—RIF算法仍具有较好的复原效果，有效地达到了保存图像细节并抑制噪声的目的。     

针对多重退化因素的印刷图像复原方法

目的对于由多种因素所导致的印刷图像退化问题,文中提出一种针对椒盐噪声、高斯噪声和模糊退化等多重退化因素的图像复原方法。方法首先针对印刷图像椒盐噪声密度不高的特点,设计一种基于灰度范围准则和局部差别准则的椒盐噪声二级检测和滤除方法,并通过评价实验得出合适的阈值参数设置。在去除高斯噪声和图像模糊的过程中,利用边缘保持平滑滤波的原理和特性,将双边滤波器和引导滤波器应用于图像复原中,又在此基础上设计和应用图像细节增强的二次引导滤波器。结果在椒盐噪声去除方面,新方法对大部分图像都能取得较好的复原效果,尤其对细微边缘不多的图像效果最佳,复原后的PSNR值能达到40以上。二次引导滤波器对高斯噪声和图像模糊的复原效果最好。结论通过对不同图像复原方法的效果进行评价和分析,验证了文中方法的性能,为今后图像复原技术的应用提供了指导。     

刑事图像技术的应用与研究

刑事图像技术是刑事科学的重要组成部分：是检验、固定痕迹物证的一种方法。刑事图像的重大应用在于扩大人类的视野，显示肉眼看不到或看不清的重要细节，以便揭露和证实犯罪。     

图像插值的改进自适应核回归方法

针对 steering 自适应核回归的插值算法中图像整体灰度值偏离较大和空洞的现象,本文提出了一种基于改进的自适应核回归函数的图像插值算法.该方法根据图像的自相似性原理,通过把图像邻域中像素间几何距离函数融入 steering 自适应核函数中,实现一种改进的自适应核函数,并采用这种核函数对图像进行回归插值.实验结果表明...     

基于多种先验结合的图像盲去模糊方法

针对传统图像盲去模糊方法中部分细节复原不准确的问题,提出了一种结合暗通道先验、梯度先验以及强度先验的图像盲去模糊方法。该方法利用清晰图像本身的稀疏性以及边缘信息的稀疏性,结合清晰图像暗通道的稀疏性,在不同权重下对清晰图像以及模糊核进行约束。通过交替迭代的方式更新潜影和模糊核,以估计出的最精细的模糊核结合非盲去模糊方法最终得到去模糊图像。实验结果表明,所提方法对图像去模糊质量有所提升,有效地恢复了图像的一些细节。     

一种椒盐噪声点的检测算法技术研究

针对中值滤波算法虽然能够在很大程度上改善噪声带来的影响,但是使图像边缘变得模糊这一问题,提出了一种检测椒盐噪声的新算法。使用该算法对图像进行噪声检测后再对噪声点进行中值滤波能够有效地去除椒盐噪声。实验证明该算法具有明显的去噪能力,且当椒盐噪声比较严重时,也能够取得较好的滤波效果。     

加权均值滤波在机场跑道路面裂纹图像检测中的应用

图像去噪是图像处理中的一个重要环节。基于均值滤波和中值滤波的经典算法,结合数字图像处理技术,以拍摄的直升机机场跑道路面裂纹图像作为研究对象,提出了一种改进的加权均值滤波算法,并通过仿真给出了试验效果图及数据结果。结果表明：改进的加权均值滤波算法较传统均值滤波能更好地保护图像的细节,失真小,在去除噪声的同时较好地保留边缘等细节信息,降低了图像处理后的模糊化程度,优于经典的滤波算法。该研究为机场跑道路面裂纹图像检测提供了一种新方法。     

Solving cell formation and task scheduling in cellular manufacturing system by discrete bacteria foraging algorithm

We consider a joint decision model of cell formation and task scheduling in cellular manufacturing system under dual-resource constrained (DRC) setting. On one hand, machines and workers are multi-functional and/or multi-skilled, and they are grouped into workstations and cells. On the other hand, there is a processing sequence among operations of the parts which needs to be dispatched to the desirable workstations for processing. Inter-cell movements of parts can reduce the processing times and the makespan but will increase the inter-cell material handling costs. The objective of the problem is to minimise the material handling costs as well as the fixed and operating costs of machines and workers. Due to the NP-hardness of the problem, we propose an efficient discrete bacteria foraging algorithm (DBFA) with elaborately designed solution representation and bacteria evolution operators to solve the proposed problem. We tested our algorithm using randomly generated instances with different sizes and settings by comparing with the original bacteria foraging algorithm and a genetic algorithm. Our results show that the proposed DBFA has better performance than the two compared algorithms with the same running time.     

JPEG基线算法对BP神经网络图像压缩处理的影响

根据BP神经网络图像压缩处理中,存在对图像信息高低频部分处理质量不同和边缘效应等问题,提出了采用JPEG基线算法于BP神经网络图像压缩处理结构中,建立了该系统。并采用灰阶Lena图像进行实验,通过实验分析发现,采用这种新的结构来处理图像,不仅可以得到较大的压缩比,而且具有较好的峰值信噪比。实验结果证明这种具有自适应性的图像处理方法,不仅可行,而且能高效、稳定地重建图像。     

Effects of processing method and fiber size on the structure and properties of wood–plastic composites

The aim of this study was to understand the roles of processing method and fiber size on the properties of wood–plastic composites (WPC). Composites were manufactured using extrusion or injection molding (IM) and fibers of different fiber length-to-diameter (L/D) ratio. IM resulted in better mechanical properties and lower water absorption and swelling than extrusion. These differences can be explained by the structure and surface quality of the composites. Fiber L/D ratio had a beneficial effect on mechanical properties but resulted in decreased water absorption characteristics. These results allow identifying a suitable forming process and/or fiber size according to application needs. This study provides a better understanding of the relationships between processing method, fiber characteristics, and composite behavior.     

固白剂在黑白照相纸中的应用试验

提高黑白照相纸影像画面白度的主要方法之一是荧光增白剂的采用，但由于荧光增白剂分子易受加工过程中的酸碱和漂洗的影响而流失，从而降低了影像画面的白度。固白剂的应用有助于提高荧光增白剂的耐光性、耐酸碱和耐漂水性能，有效地防止荧光增白剂分子在冲洗加工中的扩散流失，达到了固白增白的效果。     

快速响应光阀的制备及在真三维立体显示中的应用

固态体积式真三维立体显示系统主要由数据传输,图像处理模块,高速彩色投影光路模块,显示模块(显示体)以及外围控制电路组成。显示模块由多层液晶光阀组成,光线透过显示体后能量衰减严重,造成显示画面黯淡,模糊,此外,由于光阀响应时间慢,系统刷新频率低,屏幕存在严重的闪烁问题。本文通过在液晶中添加少量聚合物和手性物添加剂制备出快速响应的聚合物稳定胆甾相液晶光阀,其开态透过率达到88%,响应时间小于1ms。将其应用于真三维立体显示系统,显示画面清晰稳定,无闪烁。     

Channel coordination in an assembly system facing uncertain demand with synchronized processing time and delivery quantity

Supply chains need to be synchronized to increase system efficiency. With increasing globalization, an assembly or a manufacturing firm may receive supplies from all over the world. Synchronization of such a system becomes difficult for two reasons: first, each supplier has certain capacity and can supply materials with a fixed but different order processing time; second, each supplier has a different cost structure and, therefore, production quantities differ due to incentives conflicts, which leads to local optimization of profits by the suppliers. In the existing literature, this problem is generally tackled by focusing either on the synchronization of order processing time or on order quantity determination. However, when system coordination is of the essence, simultaneous consideration of both becomes imperative. A model to synchronize the assembly process is proposed and the optimal order quantities in a two-echelon assembly system with one assembler (facing a stochastic demand) and multiple suppliers with deterministic order processing times is determined. A numerical study conducted shows that risk sharing and proper safety stock placement lead to better system coordination and improve system performance.     

The effect of work content imbalance and its interaction with scheduling method on sequential flow line performance

Uneven workflow in serial flow lines is common when job work content varies. Unfortunately, work content imbalance between stations or between jobs degrades system performance. The effect of this imbalance can be mitigated by proper job sequencing. In this paper, we uncouple the influence of station-to-station and job-to-job imbalance within a set of jobs, introducing metrics to quantify the degree of imbalance, and determine their impact on line performance. We then examine the sensitivity of several scheduling methods at different problem scales and processing time variance levels, and compare the sensitivity to process time stability radii. Within the case studies explored, we determine that greater station imbalance can degrade solution quality regardless of problem scale and variance level. Greater job imbalance can either improve or diminish solution quality (i.e. relative performance from a given scheduling method), but less than station imbalance. Scheduling methods that sequence jobs based on the sums of processing times, e.g. SPT, can benefit from greater station imbalance. Scheduling methods that sequence jobs based on idle time between stations, e.g. the state-space heuristic, are more sensitive to station imbalance, but can also benefit from job imbalance. Finally, we conclude that job sequence has a greater impact on solution quality than stability radii.     

Transients in reconfigurable digital signal processing systems

To solve measurement and control problems, the processing of input data is performed typically by model-based digital signal processing (DSP) systems, which contain a representation of our knowledge about the nature and the actual circumstances of the problem at hand. If the nature or the actual circumstances change, the corresponding model should also be changed. Similarly, if the amount of knowledge about the problem increases due to measurements, an improved model can be suggested which provides better performance. As a consequence, the real-time adaptation or reconfiguration of the DSP system to be applied can hardly be avoided. The transients caused by these adaptations/reconfigurations is investigated. It is shown that for feedback systems, i.e., for infinite impulse response (IIR) filters, these transients are strongly structure-dependent and that the so-called orthogonal filter structures also provide good performance in this respect