排序方式: 共有35条查询结果,搜索用时 31 毫秒
1.
在分析比较经验模态分解(EMD)、小波变换(Wavelet)和独立分量分析(ICA)优缺点的基础上,提出一种新的EMD-Wavelet-ICA耦合模型。该模型充分利用了EMD的自适应性,对原始信号进行分解获得不同频率的模态函数(IMF),采用标准化模量的累计均值对IMF进行尺度划分;进而分别采用Wavelet和ICA对高频和低频IMF进行降噪,将降噪后的IMF进行多尺度重构,获得降噪后的信号;采用信噪比、标准差、偏差和相关系数等指标对降噪效果进行评价。仿真数据和GPS坐标序列的处理结果表明:与EMD模型和EMD-ICA模型相比,新模型的标准差、偏差均有不同程度的减小;信噪比和相关系数有一定程度的增大,可以获得更好的降噪效果。 相似文献
2.
针对传统去噪算法在复杂噪声污染图像处理中能力较弱的问题,该文基于信号高阶统计量的独立分量分析,通过其自适应变换,分离出源信号中的统计独立的分量,在分析对比传统图像去噪方法的基础上,讨论了独立分量分析的基本模型及原理;提出了一种结合中值滤波与wiener滤波的最大似然估计的图像去噪改进方法。仿真实验表明改进的独立分量分析去噪方法具有较大的优越性。 相似文献
3.
4.
针对探地雷达(Ground Penetrating Radar,GPR)信号信噪比低、背景杂波强,事先对探测目标的信息所知甚少,近乎处于“盲”状态,因而实际处理难度大等实际问题,提出了将独立分量分析(In-dependent Component Analysis,ICA)这种盲信号处理技术应用于GPR信号处理,并利用ICA中的Fast-ICA算法,对ICA法在GPR信号处理中的应用进行了初步探索,实现了GPR信号中弱目标信号和强背景杂波的有效分离,并初步解决了对所分离目标信号的正确排序,以及由ICA方法本身带来的所分离目标独立分量信号符号的不确定性等问题,使GPR信号信噪比大幅提高,从而使GPR的目标检测性能也得以显著改善。在对诸如地雷、地下管线等局部目标的时域有限差分法(Finite-dif-ference Time-domain,FDTD)、仿真GPR数据和室外试验观测GPR数据进行ICA处理后,都取得了理想的结果。 相似文献
5.
6.
郭学兰;杨敏华;毛军;周秋琳 《东北测绘》2013,(4):144-146,149+152
针对高光谱影像数据具有波段众多、数据量较大的特点,本文提出了一种基于波段子集的独立分量分析(ICA)特征提取的高光谱遥感影像分类的新方法。以北京昌平小汤山地区的高光谱影像为例,根据高光谱遥感影像的相邻波段的相关性进行子空间划分,在各个波段子集上采用ICA算法进行特征提取,将各个子空间提取的特征合并组成特征向量,采用支持向量机(SVM)分类器进行分类。结果表明:该方法分类精度最佳(分类精度89.04%,Kappa系数0.8605,明显优于其它特征提取方法的SVM分类,有效地提高了高光谱数据的分类精度。 相似文献
7.
8.
9.
10.
Landsat series multispectral remote sensing imagery has gained increasing attention in providing solutions to environmental problems such as land degradation which exacerbate soil erosion and landslide disasters in the case of rainfall events. Multispectral data has facilitated the mapping of soils, land-cover and structural geology, all of which are factors affecting landslide occurrence. The main aim of this research was to develop a methodology to visualize and map past landslides as well as identify land degradation effects through soil erosion and land-use using remote sensing techniques in the central region of Kenya. The study area has rugged terrain and rainfall has been the main source of landslide trigger. The methodology comprised visualizing landslide scars using a False Colour Composite (FCC) and mapping soil erodibility using FCC components applying expert based classification. The components of the FCC were: the first independent component (IC1), Principal Component (PC) with most geological information, and a Normalised Difference Index (NDI) involving Landsat TM/ETM+ band 7 and 3.The FCC components formed the inputs for knowledge-based classification with the following 13 classes: runoff, extreme erosions, other erosions, landslide areas, highly erodible, stable, exposed volcanic rocks, agriculture, green forest, new forest regrowth areas, clear, turbid and salty water. Validation of the mapped landslide areas with field GPS locations of landslide affected areas showed that 66% of the points coincided well with landslide areas mapped in the year 2000. The classification maps showed landslide areas on the steep ridge faces, other erosions in agricultural areas, highly erodible zones being already weathered rocks, while runoff were mainly fluvial deposits. Thus, landuse and rainfall processes play a major role in inducing landslides in the study area. 相似文献