共查询到10条相似文献,搜索用时 140 毫秒
1.
小波包熵和Fisher判别在近红外光谱法鉴别中药大黄真伪中的应用 总被引:7,自引:0,他引:7
利用傅里叶变换近红外光谱仪采集了中药大黄的近红外漫反射光谱,提取光谱的主成分和小波包熵等特征信息,再以特征信息为依据,利用Fisher分类器对中药大黄的真伪进行了鉴别。通过比较得出:采用小波包熵特征信息建模和预测误判率比采用主成分低。用小波包熵进行特征提取和Fisher分类器相结合对中药大黄真伪进行鉴别,其建模集交叉验证的误判率为6.52%,预测集的误判率是2.04%,为中药大黄的近红外快速真伪鉴别提供了参考。 相似文献
2.
针对多类运动想象情况下存在的脑电信号识别正确率比较低的问题,提出了一种将小波包方差,小波包熵和共同空间模式相结合的脑电信号特征提取,输入到支持向量机达到分类目的。首先选择小波包去噪后重要导联的脑电信号,进行小波包分解;然后对通道优化选取的重要导联的每个通道信号计算方差和熵值,对重要导联的每个通道信号的子带系数进行重构后,进行共同空间模式特征提取;最后结合2种不同导联方式所获取的特征向量进行分类。采用BCI2005desc_IIIa中l1b数据,该算法的分类正确率最高达到88.75%,相对2种单一的提取方法分别提高28.27%和6.55%。结果表明该算法能够有效提取特征向量,进而改善多类识别正确率较低的问题。 相似文献
3.
4.
Based on the techniques of Hilbert–Huang transform (HHT) and support vector machine (SVM), a noise-based intelligent method for engine fault diagnosis (EFD), so-called HHT–SVM model, is developed in this paper. The noises of a sample engine under normal and several fault states are first measured and denoised by using the wavelet packet threshold method to initially lower the noise level with negligible signal distortion. To extract fault features of the engine, then, the HHT is selected and applied to the measured noise signals. A nine-dimensional vector, which consists of seven intrinsic mode functions (IMFs) from the empirical mode decomposition (EMD), maximum value of HHT marginal spectrum and its corresponding frequency component, is specified to represent each engine fault feature. Finally, an optimal SVM model is established and trained for engine failure classification by using the fault feature vectors of the noise signals. Cross-validation results show that the proposed noise-based HHT–SVM method is accurate and effective for engine fault diagnosis. Due to outstanding time–frequency characteristics and pattern recognition capacity of the HHT and SVM, the newly proposed HHT–SVM can be used to deal with both the stationary and nonstationary signals, and even the transient ones. In the view of applications, the HHT–SVM technique may be suggested not only to detect the abnormal states of vehicle engines, but also to be extended to other fields for failure diagnosis in engineering. 相似文献
5.
Development of an algorithm for automatic detection and rating of squeak and rattle events 总被引:1,自引:0,他引:1
Unnikrishnan Kuttan Chandrika 《Journal of sound and vibration》2010,329(21):4567-4577
A new algorithm for automatic detection and rating of squeak and rattle (S&R) events was developed. The algorithm utilizes the perceived transient loudness (PTL) that approximates the human perception of a transient noise. At first, instantaneous specific loudness time histories are calculated over 1-24 bark range by applying the analytic wavelet transform and Zwicker loudness transform to the recorded noise. Transient specific loudness time histories are then obtained by removing estimated contributions of the background noise from instantaneous specific loudness time histories. These transient specific loudness time histories are summed to obtain the transient loudness time history. Finally, the PTL time history is obtained by applying Glasberg and Moore temporal integration to the transient loudness time history. Detection of S&R events utilizes the PTL time history obtained by summing only 18-24 barks components to take advantage of high signal-to-noise ratio in the high frequency range. A S&R event is identified when the value of the PTL time history exceeds the detection threshold pre-determined by a jury test. The maximum value of the PTL time history is used for rating of S&R events. Another jury test showed that the method performs much better if the PTL time history obtained by summing all frequency components is used. Therefore, rating of S&R events utilizes this modified PTL time history. Two additional jury tests were conducted to validate the developed detection and rating methods. The algorithm developed in this work will enable automatic detection and rating of S&R events with good accuracy and minimum possibility of false alarm. 相似文献
6.
在八种化学计量学除噪技术比较研究的基础上,研制了小波包变换Elman回归神经网络方法(WPERNN)用于研究重叠光谱的同时定量测定。结合小波包变换和Elman回归神经网络改进除噪质量及回归能力。通过最佳化,选择了小波函数、小波包分解水平和Elman回归神经网络(ERNN)的结构及参数。两个程序PWPERNN和PERNN被设计执行WPERNN和ERNN方法计算。七种化学计量学方法用于比较研究。实验结果显示WPERNN方法是成功的且优于其他六种方法。 相似文献
7.
基于小波包和偏最小二乘支持向量机的多光谱纹理图像的大米分类研究 总被引:4,自引:2,他引:2
提出了一种利用多光谱图像纹理特征进行大米分类的新方法.图像由MS3100-3CCD光谱成像仪获得,光谱成像仪提供3个波段的图像,由近红外(NIR)、红色(R)和绿色(G)组成,因此它能够获取普通数码照相机所不能获取的信息.对3CCD近红外波段图像进行二层小波包分解,得到16个子频带,因为纹理图像的特征信息主要集中在中频,因此提取8个中频频带(带通频带)的熵值,并且作为支持向量机的特征值输入.最后应用支持向量机技术分别对有和没有经过小波包分解的NIR波段纹理图像的熵值进行建模,建模样本和预测模型各为80个,每种各为20个.对四种大米进行处理,结果表明,经过小波包分解的纹理图像的识别率达到了100%,而没有经过小波包分解的纹理图像的识别率只有93.75%,说明结合小波包和支持向量机进行多光谱图像的纹理识别是种非常有效的技术,同时也为大米的分类提供一种快速和无损的新方法. 相似文献
8.
9.
10.
Image encryption with fractional wavelet packet method 总被引:2,自引:0,他引:2
We introduce a new method called fractional wavelet packet transform to encrypt images in this paper, in which fractional orders and wavelet packet filter are its two series of keys. Fractional orders are additional keys in this method compared to wavelet packet encryptions. Selected image encryption is also proposed in this paper, and it is quite more flexible and effective than wavelet, fractional wavelet or wavelet packet encryptions. The possible optical implementation and digital computation are proposed. Computer simulations prove its feasibility. 相似文献