Fisher分布下基于白化滤波的极化SAR图像CFAR检测方法

在已有的极化合成孔径雷达（SAR）图像恒虚警（CFAR）检测方法中,存在着高分辨下杂波模型适用性差的难题.为此提出了一种Fisher分布下具有虚警概率解析表达形式的CFAR检测方法.首先,在乘积模型框架下,引入Fisher纹理变量,推导出了极化白化滤波（PWF）检测量的概率密度函数（PDF）.然后,对PDF积分得到了虚警概率关于检测门限的解析表达形式,并设计了相应的CFAR检测算法流程.最后,通过机载合成孔径雷达（AIRSAR）实测数据比较了新方法和双参数恒虚警（2P-CFAR）算法及已有的基于K分布、G₀分布、Wishart分布的CFAR检测方法的检测性能.结果表明新方法能有效检测出目标,且鲁棒性较强,相比于其他检测方法,品质因数平均高出32.66%.     

Selection of NF membrane to improve quality of chemically treated surface water

The requirement for higher quality drinking water necessitates the application of more efficient water treatment techniques. Nanofiltration is one promising option for enhanced water treatment, for example, in enhanced organic matter removal. The characteristics of different nanofiltration membranes vary remarkably, and the selection of a membrane has to be made according to the requirements of an application. In this study six nanofiltration membranes (NF70, NF255, NTR-7450, NTR-7410, Desal-5 and TFC-S) were evaluated in improving the quality of chemically pre-treated surface water in a pilot-scale process. The results indicate that the membrane with high organics removal and slightly reduced ion removal characteristics (NF255) performed best in terms of product water quality as well as membrane productivity and fouling. The most permeable membrane (NTR-7410) suffered intensive fouling and insufficient product water quality. An interesting finding was that the permeates of all the tested membranes possessed a significant potential for microbial growth, despite the low nutrient contents.     

Fluid dynamic gauging of microfiltration membranes fouled with sugar beet molasses

Fluid dynamic gauging (FDG) was used to track the thickness of the cake layer formed during the microfiltration of a 45° Brix molasses solution using a 1.5 μm polysulphone membrane. A simultaneous measure of flux and deposit thickness throughout the full membrane operating cycle is reported. Asymptotic fouling thicknesses of ca. 100 μm are developed after 30 min of filtration. Accordingly, flux declines are severe at ca. 93%. Permeate line closed (PLC) operation leads to the complete removal of the deposit layer, and the recovery of 60% of the flux. However, permeate line open (PLO) operation leads to only a 50% flux recovery and an asymptotic deposit thickness of 10 μm. An initial increase and subsequent reduction in flux during chemical cleaning has also been recorded for both acid and alkali cleaning regimes. Effective cleaning regimes for membranes fouled by molasses require an alkali stage followed by an acid stage.     

基于联合PWF2pCFAR算法的Envisat/ASAR双极化数据舰船检测方法研究

利用极化白化滤波PWF算法融合Envisat/ASAR双极化复数据图像,同时抑制相干斑,然后利用双参数恒虚警率2pCFAR算法进行舰船目标检测。在不同检测门限和检测模板条件下,利用实测VV/HV极化组合数据进行的仿真实验结果表明,联合PWF 2pCFAR算法不仅能够有效地检测出舰船目标,较好地降低了虚警率,也能够保留舰船目标的结构信息;同时得到了一组相对最优的检测门限与检测模板,具有一定的实际参考意义。     

极化合成孔径雷达图像船舶目标检测算法

结合区域划分和结构检测模板提出了改进极化白化滤波(IPWF)算法,利用IPWF算法融合极化合成孔径雷达(POL-SAR)中各极化通道图像,同时抑制相干斑,然后利用双参数恒虚警率(CFAR)检测方法对融合后的图像进行船舶目标检测.本文利用香港地区SIR-C全极化单视复数据进行了实验,结果表明IPWF算法更好地降低了相干斑因子,提高了船舶目标的检测率、控制了虚警率,同时可以更好地保持船舶目标的结构信息.     

基于联合PWF-2pCFAR算法的Envisat/ASAR双极化数据舰船检测方法研究

利用极化白化滤波(PWF)算法融合Envisat/ASAR双极化复数据图像,同时抑制相干斑,然后利用双参数恒虚警率(2pCFAR)算法进行舰船目标检测。在不同检测门限和检测模板条件下,利用实测VV/HV极化组合数据进行的仿真实验结果表明,联合PWF-2pCFAR算法不仅能够有效地检测出舰船目标,较好地降低了虚警率,也能够保留舰船目标的结构信息;同时得到了一组相对最优的检测门限与检测模板,具有一定的实际参考意义。     

19通道多孔陶瓷膜渗透过程的CFD模拟

根据Darcy定律，建立了多孔陶瓷膜中渗透流动的CFD计算模型，定量描述了纯水在多孔陶瓷膜中的渗流情况。根据CFD模型，采用有限体积法，模拟计算出四种平均孔径陶瓷膜的纯水通量，并与实验值进行了对比，结果吻合较好。采用该模型，定量计算出19通道陶瓷膜中每个通道对整体陶瓷膜渗透通量的贡献，并分析了其中的变化规律，为多孔陶瓷膜的构型优化奠定了基础。     

极化SAR相干斑抑制极化白化滤波器

相干斑抑制是合成孔径雷达(SAR)图像和极化SAR图像后处理的首要问题。该文针对最优的极化相干斑抑制滤波器极化白化滤波器(PWF)仅输出一个强度图像的缺陷,对 PWF进行了改进,得到了各个极化通道的强度图像。     

极化白化滤波器的一种多通道扩展

为减小相干斑对分类、检测的影响,通常要对极化合成孔径雷达图像进行相干斑抑制。极化白化滤波器(PWF)是利用多个极化通道的复数据抑制相干斑的有效方法,但其输出只有一个通道。为了在后续处理中能够利用全极化数据,该文对PWF进行了改进,提出了多通道极化白化滤波器法(MCPWF)。该方法在保持PWF最少相干斑图像输出的同时,还能得到各个极化通道滤波后的输出。Pi-SAR实测数据的实验结果验证了该方法的有效性。