搜索资源列表
A_subspace_algorithm
- 子空间算法是一种基于矩阵特征空间分解的方法,信号子空间由阵列接收到的数据协方差矩阵中与信号对应的特征向量组成噪声子空间则由协方差矩阵中所有最小特征值噪声方差对应的特征向量组成。子空间算法就是利用这两个互补空间之间的正交特性来估计空间信号的方位-Subspace algorithms is minimized by the corresponding eigenvalues of all the noise variance-covariance matrix of a
Bxiaobo3
- 小波包分解,提取能量特征,然后进行归一化处理,获取诊断信息-Wavelet Neural Network
POD
- 对数据进行本征正交分解,是分析周期性本质特征的重要算法-Data orthogonal decomposition is an important characteristic of algorithm analysis cyclical nature
LMDchengxu
- 采用小波相邻系数降噪,然后用改进的局域均值分解对滚动轴承进行特征提取,得到特征向量-Using adjacent wavelet coefficients of noise reduction, and then use the improved local mean decomposition characteristics of rolling bearing are extracted, get the feature vector
emd
- EMD经验模态分解,用于信号处理,也常和hilbert变换一起用于信号特征提取、信号消噪等领域-EMD empirical mode decomposition for signal processing, and is often used in conjunction with hilbert transform signal feature extraction, signal de-noising and other fields
matlab.QIyizhifenjie
- 关于奇异值分解,图像特征融合,图像处理的特征图提取方法-About the singular value decomposition
emd
- 可以有效方便地对原始数据进行自适应分解,广泛用于故障诊断和模式识别等方面的特征提取。-Can be effectively and easily decompose the raw data adaptively, widely used in troubleshooting and other aspects of the pattern recognition feature extraction.
dbf2
- 针对16阵元的均匀直线阵,使用MVDR算法和DREC两种方法分析方向图性能,DREC法进行特征值分解,大特征值对应的特征向量构成信号子空间。-For 16 array element uniform linear array, eigendecomposed MVDR algorithm using two methods and DREC pattern analysis performance, DREC method, large eigenvalues eigenvectors corr
baoluopu
- 对一段振动信号emd分解后,求每个分量的峭度值,选取峭度最大的分量,求其包络谱,能够明显看出特征频率。-After a period of vibration signal emd decomposition, seeking kurtosis value of each component, the largest component of kurtosis, seeking its envelope spectrum, can obviously see that the characte
HHT
- Hilbert-Huang变换(HHT)是一种新的非平稳信号处理技术,该方法由经验模态 分解(EMD)与Hilbert谱分析两部分组成。任意的非平稳信号首先经过EMD方法处理后被分解为一系列具有不同特征尺度的数据序列,每一个序列称为一个固有模态函数(IMF),然后对每个IMF分量进行Hilbert谱分析得到相应分量的Hilbert谱,汇总所有Hilbert谱就得到了原信号的谱图。该方法从本质上讲是对非平稳信号进行平稳化处理,将信号中真实存在的不同尺度波动或趋势逐级分解出来,最终用瞬时频率和能量来
lmd_outer
- 轴承外圈12K实际信号LMD分解程序,信号包络频谱,能明显提取故障特征-LMD bearing outer ring 12K actual signal decomposition process, the signal envelope spectrum, can significantly fault features
KL_SVD_face_recognition
- PCA主成分分析,采用KL投影和SVD分解提取人脸特征向量,最后采用最近邻判别法计算识别率。-Face recognition based on PCA. KL projection and SVD are used to extract face eigenvectors. Recognition rate is calculated by k nearest neighbors(KNN) method.
CHENGXU
- MUSIC算法[1] 是一种基于矩阵特征空间分解的方法。从几何角度讲,信号处理的观测空间可以分解为信号子空间和噪声子空间,显然这两个空间是正交的。信号子空间由阵列接收到的数据协方差矩阵中与信号对应的特征向量组成,噪声子空间则由协方差矩阵中所有最小特征值(噪声方差)对应的特征向量组成。MUSIC算法就是利用这两个互补空间之间的正交特性来估计空间信号的方位。噪声子空间的所有向量被用来构造谱,所有空间方位谱中的峰值位置对应信号的来波方位。MUSIC算法大大提高了测向分辨率,同时适应于任意形状的天线阵列
jdqr
- 采用QR分解方法计算矩阵特征值和特征向量-computes a partial Schur decomposition of a square matrix or operator
jdqz
- 采用部分Schur或QZ分解计算矩阵特征值和特征向量-computes a partial generalized Schur decomposition (or QZ decomposition) of a pair of square matrices or operators
emd
- EMD分解,改算法可以有效地对数据进行频率特征提取-EMD decomposition algorithm can change the frequency characteristics of the data extraction
tuxiangtezhengtiqu
- 对图像的特征提取,有灰度直方图,共生矩阵,Hu不变矩,几何特征,还有基于小波的二维分解,已运行。-Image feature extraction, there histogram, co-occurrence matrix, Hu invariant moments, the geometrical characteristics, already running, correct.
NMF_picture
- 非负矩阵分解,对原始图像进行特征提取,也可用于去噪等。-nonegative matrix factorization
orldata-and-ONMF
- 正交非负矩阵分解,适合做非负矩阵分解的用,能得到更稀疏和更正交的基特征-Orthogonal nonnegative matrix factorization,Suitable for use NMF, can be more sparse and correct post basis vectors
WAVELET-FENJIE
- 小波变换分解与重构,选取合适频率的信号重构特征信号-Wavelet transform decomposition and reconstruction, the appropriate signal frequency characteristic signal reconstruction