梯度相位检测在医学图像小波边缘提取中的应用

针对小波变换不能涉及过多尺度的问题,分析了边缘类型对小波多尺度边缘提取的影响,并结合医学图像的特点,提出基于Canny算子的梯度相位法。梯度相位法认为边缘不只是在灰度发生突变的地方存在,如果某个区域,灰度沿某个方向缓慢变化。这个区域也存在边缘,只不过可以认为边缘比较粗。通过对人脑后颅骨的CT医学图像边缘提取的实验及结果定量和定性分析,证明此方法在医学罔像边缘检测中是切实可行的。     

医学图像的多进制小波分解

主要介绍了多进制小波基本理论并分析了其变换特征,对医学图像做了八进制小波分解,它在医学图像处理中会有更大的应用前景.     

板坯边角裂纹的小波检测方法

提出一种通过小波变换检测板坯边角裂纹的方法。该方法针对边角裂纹图像的特点,利用小波对其进行多尺度分析,在检测出边缘线的同时将图像分解为4个方向的细节分量;再选择恰当的阈值对垂直分量图进行分割．得到最终的检测结果。实验表明：与其他两种常用检测方法相比,该方法能更有效地提取出板坯边角裂纹。     

基于多尺度小波残差网络的稀疏角度CT图像恢复

目的 稀疏角度CT具有加速数据采集和减少辐射剂量的优点。然而,由于采集信息的减少,使用传统滤波反投影算法（FBP）进行重建得到的图像中伴有严重的条形伪影和噪声。针对这一问题,本文提出基于多尺度小波残差网络（MWResNet）对稀疏角度CT图像进行恢复。方法 本网络中将小波网络与残差块相结合,用以增强网络对图像特征的提取能力和加快网络训练效率。实验中使用真实的螺旋几何CT图像数据“Low-dose CT Grand Challenge”数据集训练网络。通过观察图像表征和计算定量参数的方法对结果进行评估,并与其他现有网络进行比较,包括图像恢复迭代残差卷积网络（IRLNet）,残差编码解码卷积神经网络（REDCNN）和FBP卷积神经网络（FBPConvNet）。结果 实验结果表明,本文提出的多尺度小波残差网络优于其余对比方法。结论 本文提出的MWResNet网络能够在保持稀疏角度CT图像边缘细节信息的同时有效抑制噪声和伪影。     

强噪声条件下基于小波包能量的语音增强新算法

目的：研究在强噪声背景条件下增强语音质量的方法,为在复杂条件下获取语音信号奠定基础.方法：在应用小波包分析技术对语音信号进行分解与重构的基础上,对分解后的小波包系数进行尺度,时间2个方面的阈值自适应调节,再对此系数进行重构以实现语音信号的噪声自适应消除.结果：在信噪比为0dB的强噪声条件下,在0～3000Hz较宽的频率段上,增强后的语音频谱明显清晰,且各频谱成份更加丰富.结论：本方法能够在强背景噪声条件下对语音信号中的噪声成分进行有效去除.     

基于小波包分解和 HMM模型的纹理分析

本文用小波包分解法 (WPD)对肝脏 B超图像的分类进行了研究,分类对象为正常肝图像和脂肪肝图像二类,这些图像分近程图像和远程图像来分别对待.用隐含马尔可夫模型 (HMM)分类.实验结果显示该法分类正确率要比多分辨分形特征法 (MFF)高,是一种潜在的分析 B超肝脏图像纹理的工具.     

基于边缘检测的CT图像粗糙集增强算法

粗糙集理论是一种处理模糊和不确定性问题的工具,是一种非线性处理方法.本文基于边缘检测对图像进行小波阈值去噪,在此基础之上引入粗糙集理论划分子图并分别进行增强处理,再进行对比度增强.该算法把边缘检测去噪与粗糙集理论结合起来增强图像,实践证明具有良好的图像增强效果.     

多尺度增强算法在肺结节计算机辅助检测中的应用探讨

目的研究多尺度增强算法对肺小结节的增强能力及将其作为肺结节计算机辅助检测的处理方法的可行性。方法针对肺结节的形态特点,采用高斯函数模拟肺结节,运用多尺度图像增强滤波器算法,增强肺结节提取兴趣区,供后续的分类判别使用。结果通过对肺部CT图像的应用,说明多尺度增强算法能很好地检测疑似肺结节兴趣区。结论多尺度增强算法对帮助医师检测肺结节有明显的作用,是一种有效的图像预处理方法,对肺结节的计算机辅助检测有较大应用价值。     

Snake算法边缘检测图像分析

随着图像信息应用需要的迅速增长,数字图像处理技术在现今社会中有着越来越多的应用,特别是边缘检测技术.本文探讨了边缘检测技术中Snake算法的原理及方法,编程实现了文中所提出的检测方法,并对图像进行了实际处理,取得了较好的效果.     

一种基于色差的彩色图像的边缘检测方法

传统的边缘检测方法大都是基于灰度图像的,这就不能充分利用彩色图像的全部信息。文章提出了一种基于色差的彩色图像的边缘检测方法,它能快速准确地检测到图像的边缘,具有低的错判率和较高的效率,优于传统基于灰度的边缘检测方法。     

A New Approach to Detection of ECG Arrhythmias: Complex Discrete Wavelet Transform Based Complex Valued Artificial Neural Network

This paper presents the new automated detection method for electrocardiogram (ECG) arrhythmias. The detection system is implemented with integration of complex valued feature extraction and classification parts. In feature extraction phase of proposed method, the feature values for each arrhythmia are extracted using complex discrete wavelet transform (CWT). The aim of using CWT is to compress data and to reduce training time of network without decreasing accuracy rate. Obtained complex valued features are used as input to the complex valued artificial neural network (CVANN) for classification of ECG arrhythmias. Ten types of the ECG arrhythmias used in this study were selected from MIT-BIH ECG Arrhythmias Database. Two different classification tasks were performed by the proposed method. In first classification task (CT-1), whether CWT-CVANN can distinguish ECG arrhythmia from normal sinus rhythm was examined one by one. For this purpose, nine classifiers were improved and executed in CT-1. Second classification task (CT-2) was to recognize ten different ECG arrhythmias by one complex valued classifier with ten outputs. Training and test sets were formed by mixing the arrhythmias in a certain order. Accuracy rates were obtained as 99.8% (averaged) and 99.2% for the first and second classification tasks, respectively. All arrhythmias in training and test phases were classified correctly for both of the classification tasks.     

基于Canny算子的医用X光图像边缘检测算法研究

为了进一步加强数字图像处理技术在医学图像领域的应用,提供一种基 Canny算子的医用X先图像边缘检测方法.首先介绍边缘检测算法的理论基础,然后结合Canny算子的实现过程,说明边缘检测的所遵循的基本原则,最后对该算法进行了实验验证.从实验的结果可以看出,这种方法对医用X光图像非常有效,对利用计算机进行图像处理有着重要的现实意义.     

基于小波包分析和SVM的在线手写签名鉴别

针对在线手写签名难以提取有效特征的实际情况,提出用小波包分解和单支重构来构造能量特征向量的方法,直接利用各频段成分能量的变化来反映签名的动态特征。给出了衡量各特征识别能力的Fisher准则,并且基于该准则剔除了识别能力差的特征,优化了特征空间。用该方法构造的特征向量能突出反映签名的动态特征。然后采用SVM对签名进行识别。实验证明:采用本文方法识别的正确率高达99.38%,错误拒绝率FRR=0.25%,错误接受率FAR=1.0%,其性能令人满意。     

A RONI Based Visible Watermarking Approach for Medical Image Authentication

Nowadays medical data in terms of image files are often exchanged between different hospitals for use in telemedicine and diagnosis. Visible watermarking being extensively used for Intellectual Property identification of such medical images, leads to serious issues if failed to identify proper regions for watermark insertion. In this paper, the Region of Non-Interest (RONI) based visible watermarking for medical image authentication is proposed. In this technique, to RONI of the cover medical image is first identified using Human Visual System (HVS) model. Later, watermark logo is visibly inserted into RONI of the cover medical image to get watermarked medical image. Finally, the watermarked medical image is compared with the original medical image for measurement of imperceptibility and authenticity of proposed scheme. The experimental results showed that this proposed scheme reduces the computational complexity and improves the PSNR when compared to many existing schemes.     

A New Approach for Concealed Information Identification Based on ERP Assessment

Recently, numerous concealed information test (CIT) studies have been done with event related potential (ERP) for its sufficient validity in applied use. In this study, a new approach based on wavelet coefficients (WCs) and kernel learning algorithm is proposed to identify concealed information. Totally 16 subjects went through the designed CIT paradigm and the multichannel electroencephalogram (EEG) signals were recorded. Then, the high-dimensional WCs of ERP in delta, theta, alpha and beta rhythms were extracted. For the analysis of the data, kernel principle component analysis (KPCA) and a support vector machines (SVM) classifier are implemented. The results show that WCs features are significant differences between concealed information and irrelevant information (P?<?0.05). The KPCA is able to effectively reduce feature dimensionalities and increase generalization performance of SVM. A high accuracy (93.6%) in recognizing concealed information and irrelevant information is achieved, which indicates the combination KPCA and SVM may provide a useful tool for detecting the concealed information.     

基于多尺度熵和支持向量机的人体脑疲劳状态检测研究

目的探索多尺度熵结合支持向量机的方法是否可以有效检测人体脑疲劳状态,进而比较不同脑皮层位置的电极检测脑疲劳的效果。方法通过持续认知负荷任务建立脑疲劳模型,采用一款便携式脑电设备采集12名实验对象清醒和疲劳状态的脑电信号,以多尺度熵为特征,结合支持向量机算法对两种状态的脑电进行分类。结果在进行持续认知负荷任务后,实验对象的疲劳程度明显上升,NASA-TLX和KSS量表结果均具有显著的统计学差异(P<0.01);在额叶Fpz、顶叶Pz和枕叶Oz三个电极,实验对象清醒和疲劳状态脑电信号的平均分类准确率分别为92.16%、81.63%和90.54%,其中Fpz和Oz电极之间没有统计学差异(P>0.05),二者和Pz电极之间有显著的统计学差异(P<0.05)。结论多尺度熵结合支持向量机可以有效地对人体清醒和脑疲劳状态进行检测,Fpz和Oz电极比Pz电极的检测效果更好。     

A New Method for Computed Tomography Angiography (CTA) Imaging via Wavelet Decomposition-Dependented Edge Matching Interpolation

The interpolation technique of computed tomography angiography (CTA) image provides the ability for 3D reconstruction, as well as reduces the detect cost and the amount of radiation. However, most of the image interpolation algorithms cannot take the automation and accuracy into account. This study provides a new edge matching interpolation algorithm based on wavelet decomposition of CTA. It includes mark, scale and calculation (MSC). Combining the real clinical image data, this study mainly introduces how to search for proportional factor and use the root mean square operator to find a mean value. Furthermore, we re- synthesize the high frequency and low frequency parts of the processed image by wavelet inverse operation, and get the final interpolation image. MSC can make up for the shortage of the conventional Computed Tomography (CT) and Magnetic Resonance Imaging(MRI) examination. The radiation absorption and the time to check through the proposed synthesized image were significantly reduced. In clinical application, it can help doctor to find hidden lesions in time. Simultaneously, the patients get less economic burden as well as less radiation exposure absorbed.     

QT间期测量的小波分析方法

目的　准确获取人体心电信号 (ECG)中的 QT间期值。方法　利用小波变换多尺度多分辨的特点 ,将心电信号进行多尺度分解 ,把不同频带的信号显现在小波分解各个尺度上。特征尺度上准确定位 QRS波及T波的起始点 ,从而获得 QT间期的精确值。结果　 1在 2 m V的心电信号上引入峰峰值为 1m V的误差信号 ,利用二进小波进行分解和重构得到相对精确的心电波形 ,测得 QRS波、P波、T波的宽度最大误差分别为 5 .75 % ,5 .2 % ,3.8%。 2用小波分析方法对 MIT/ BIH数据库中的 ECG信号进行处理后 ,测量得到各间期的平均值及标准偏差 ,可以反映出 QT间期及相应各间期的稳定性。结论　人体心电信号随着检测状态及时间的变化具有明显的非平稳性及包含许多干扰的特点。利用小波变换将心电信号进行处理能够获得 QT间期的精确值 ,为临床诊断提供了更加准确的依据