Description: 提出了利用小波分解建立多分辨率图像锥和Hausdorff距离的医学图像配准方法。先利用小波方法建立多分辨率图像锥, 然后根据梯度向量幅度提取分层图像的特征点, 利用Hausdorff距离进行特征点集的匹配。该方法提高了配准的速度和精度, 而且具有鲁棒性。-Image registration is the matching processing in which two or more images match from the same scene derived from different time and different sensors. A new method of medical image registration based on multi-resolution image cone and Hausdorff Distance is put forward. Firstly, the image is decomposed with wavelet method and transformed to multiresolution image. Multiresolution image cone is constructed by multiresolution image. Secondly, gradient vector flow is used to detect feature points. Finally, Hausdorff Distance is used for matching sets of feature points. Theories and experiments indicate that this method has some advantages such as high precision and good robustness. Platform: |
Size: 300032 |
Author:vivi |
Hits:
Description: 电压跌落是最严重的动态电能质量问题之一，
精确定位电压跌落起止时间是应对电压跌落问题的
重要前提和基础。由于电压采样信号往往有噪声分
量，现有的方法在定位电压跌落的起止时间时存在
局限性。本文提出利用多小波变换及相邻系数去噪
的电压跌落定位方法。多小波兼有对称性、正交性、
有限支撑性和二阶消失矩等优异的信号处理性能，
利用GHM多小波可以准确定位电压跌落起止时间。
多小波变换系数在每层之间具有对应关系，多小波
相邻系数将紧相邻的若干个系数作为一个整体来确
定阈值，考虑了系数之间的相关性，能获得更好的
去噪效果。通过 Matlab 进行仿真验证，仿真结果表
明，所提出的方法的正确性。
-Voltage sag is one of the most serious
dynamic power quality problems. Critical start-time
and end-time are important indices for voltage sags.
But the sampling signals often have noisy component,
the locations of start-time and end-time are hard to get.
Wavelet is an effective tool for those non-stationary
signal processing and has been used in this field. Local
feature in the signal can be enlarged after the
transformation using the scalar wavelet. But scalar
wavelets cannot contain orthogonality, symmetry,
compact support and higher order of vanishing
moments simultaneously. In this thesis, multi-wavelets
GHM is used to detect and locate power quality
disturbances. Multi-wavelets offer many excellent
properties such as the same approximation order but
more compact support. The dependence of the
multi-wavelets coefficients varies with the level, so
neighboring coefficien Platform: |
Size: 1363968 |
Author:李荣 |
Hits:
Description: In this paper, the expert system is introduced in
order to detect and classify commonly power quality
disturbances. This system is using learning vector quantization
artificial neural networks. Clustering method named fuzzy
c-mean is also utilized to initialize weight vector of first hidden
layer. It can mitigate the disadvantage of LVQ ANN. The
proposed system employs wavelet decomposition coefficients for
extracting of deviated signals features. The determined feature
vector is derived from Standard Deviation of 10-level
decomposition detail coefficients. For the purpose of having
efficient network, just 3 characteristic points among 10 points
have been used, that leads to make networks training much Platform: |
Size: 4096 |
Author:applepie12356 |
Hits:
Search - wavelet feature detect