An indexed atlas of digital mammograms for computer-aided diagnosis of breast cancer

Computer-aided diagnosis (cad) systems are being developed to assist radiologists in the interpretation of ambiguous mammographic features corresponding to possible signs of early breast cancer. Databases of digital mammograms are needed for testing such systems; we present an overview of a few such databases. Most databases are limited to single-exam sets of two or four mammograms on which the diagnosis was made, some ground-truth information related to the position of diagnostically significant mammographic features, and the diagnosis. We propose the design of a comprehensive, indexed atlas of digital mammograms. The design of an appropriate indexing scheme facilitates the implementation of content-based retrieval techniques needed for efficient access to and retrieval of relevant cases from the atlas. We also propose the use of mobile software agents for facilitating remote consultation of the atlas. Mobile agents can move between data sources such as the atlas and hospital repositories, perform computational tasks at each site, and return only relevant data to the user. These features reduce the computational requirements of the local computer system, bandwidth requirements, and overall network traffic. Proposed applications of the atlas include research, remote consultation, teaching, evaluation ofcad systems, and self-evaluation by radiologists.     

基于SVM与SNMP的TRDP网络故障诊断方法研究

列车通信网络是列车的中枢神经系统,而列车实时数据协议TRDP(Train Real-time Data Protocol)是列车通信网络能否高效可靠运行的保障。文中提出一种基于简单网络管理协议SNMP(Simple Network Management Protocol)的网络特征提取方法和基于支持向量机SVM(Support Vector Machine)的TRDP网络故障诊断方法。实验结果表明,MIB(Management Information Base)节点的网络统计数据可以表征网络状态,基于SVM的故障诊断方法可以有效地诊断网络故障。     

基于Gabor的特征提取算法在人脸识别中的应用

针对人脸识别中的特征提取问题,提出一种新的基于Gabor的特征提取算法,利用Gabor小波变换良好的提取区分能力和LDA所具有的判别性优势来进行特征提取。首先利用Gabor小波变换来提取人脸特征。然后对得到的高维特征采用PCA进行初次降维,再利用LDA实现再次降维,得到最终的特征向量。在ORL和YALE人脸库上的实验验证了该算法的有效性。     

Editorial: Signal processing techniques for ISAR and feature extraction

Inverse synthetic aperture radar (ISAR) is a non-cooperative target recognition technique that has been investigated for target identification by the combat identification (ID) community for the past decade. Recently, ISAR imaging of moving targets has been an area of vigorous research. ISAR imaging is an effective way to acquire high resolution images of targets of interest at long range and as such is an irreplaceable tool in the task of non-cooperative target recognition. Its applications include detection, imaging, and classification of ships and aircraft with airborne, maritime, and land-based radar systems. Being radar-based, this imaging technique can be employed in all weather and day/night conditions.     

基于自适应特征提取的数显仪表识别系统

为了适应数显仪表各种不同的室外环境,建立了自适应于不同光照条件下的仪表自动识别系统。针对数显仪表图像的点亮区域多为红色,不亮部分偏灰白的特征,设计基于RGB彩色空间滤波方法,较好地保留了点亮区域的亮度信息;并提出改进的最大类间方差法,实现了数显仪表图像字符的自适应提取。根据数显仪表数字特征信息,采用穿线识别法,实现了数显仪表的自动识别。通过大量实测实验表明,提出的自适应特征提取算法对不同角度、不同光照环境拍摄的仪表图像都有良好的适应性,应用于数显仪表自动识别系统可得到95%的识别准确率,平均运行时间为1.136 s。     

Computerized analysis of multiple-mammographic views: potentialusefulness of special view mammograms in computer-aided diagnosis

PURPOSE: To investigate the potential usefulness of special view mammograms in the computer-aided diagnosis of mammographic breast lesions. MATERIALS AND METHODS: Previously, we developed a computerized method for the classification of mammographic mass lesions on standard-view mammograms, i.e., mediolateral oblique (MLO) view and/or cranial caudal (CC) views. In this study, we evaluate the performance of our computerized classification method on an independent database consisting of 70 cases (33 malignant and 37 benign cases), each having CC, MLO, and special view mammograms (spot compression or spot compression magnification views). The mass lesion identified in each of the three mammographic views was analyzed using our previously developed and trained computerized classification method. Performance in the task of distinguishing between malignant and benign lesions was evaluated using receiver operating characteristic analysis. On this independent database, we compared the performance of individual computer-extracted mammographic features, as well as the computer-estimated likelihood of malignancy, for the standard and special views. RESULTS: Computerized analysis of special view mammograms alone in the task of distinguishing between malignant and benign lesions yielded an Az of 0.95, which is significantly higher (p < 0.005) than that obtained from the MLO and CC views (Az values of 0.78 and 0.75, respectively). Use of only the special views correctly classified 19 of 33 benign cases (a specificity of 58%) at 100% sensitivity, whereas use of the CC and MLO views alone correctly classified 4 and 8 of 33 benign cases (specificities of 12% and 24%, respectively). In addition, we found that the average computer output of the three views (Az of 0.95) yielded a significantly better performance than did the maximum computer output from the mammographic views. CONCLUSIONS: Computerized analysis of special view mammograms provides an improved prediction of the benign versus malignant status of mammographic mass lesions.     

小波变换在光谱特征提取方面的应用

人们在处理高光谱图像时一般要对一些典型地物进行光谱分析、特征波段的提取,以便提取出最大量的有效信息,剔除无用或冗余的信息,然后再进行分类识别.采用小波变换的分析方法,选用合适的小波进行分解,根据分解后的高频分量中包含的重要信息,利用局部相邻的正负极值点找出对应于原始光谱曲线上每个吸收带的左右边界;利用局部过零点,即可比较精确的提取出各个吸收带的中心波长.该方法比传统的光谱特征提取方法更简洁、有效,实验证明为一种比较理想的光谱特征提取方法.     

基于环形核的旋转不变性特征提取方法

针对人脸在XOZ平面内旋转,即在图像所在平面内人脸产生的旋转,因特征值变化大导致人脸识别率降低的问题,提出了一种新颖的基于环形核的旋转不变性特征(Circular Kernel Feature,CKF)提取方法.所提算法有两个创新点,第一点是给出了环形核的建立方式,定位人脸上明显的特征部分.第二点是提供了特征的旋转不变计算方式.首先建立环形核,定位人脸上明显特征的坐标区域;然后,用旋转不变的计算方式获取定位区域的特征值.在Georgia Tech人脸数据库上的实验证明:人脸旋转前后CKF的值相较Gabor,LBP等特征值的变化小了98％.     

通信辐射源个体特征提取技术

个体特征选择和提取是辐射源个体识别的关键,直接决定分类识别性能的好坏。由于在实际工程应用中,利用暂态特征进行通信辐射源个体识别难以实现,本文从稳态特征出发,对通信辐射源个体特征提取技术进行了综述,对特征的产生机理、在信号传播过程中所受到的污染以及在实际工程应用中的可行性做了归纳与分析。最后,指出了目前通信辐射源个体特征提取技术存在的问题,展望了个体识别技术未来可能的研究方面。     

基于FPGA的数字水印提取系统设计研究

针对信息安全问题的日益突出,提出了基于5/3整数小波的数字水印算法,给出了两种5/3小波的硬件架构:一种是基于RAM的流水线架构;另一种是基于行分组的行列并行架构。进而设计了基于FPGA的数字水印提取硬件系统,结果证明该算法具有很好的不可见性及鲁棒性,且复杂度低,硬件较容易实现,并将水印提取代码下载到FPGA硬件进行验证,结果证明该算法可以很好地实现水印的提取工作。     

基于Matlab GUI的服装颜色特征提取系统设计

针对传统服装颜色特征提取系统直观性差,提取结果不可靠等问题,设计了基于Matlab GUI的服装颜色特征提取系统。介绍了系统设计思想,对服装图片进行主成分分析变换,得到主成分分析空间的颜色特征,把测试样本和训练样本依次投影至特征空间,得到相应的识别特征,采用K-means法对颜色进行分类,按照空间一致原则实现服装颜色特征提取。给出GUI界面设计过程,采用Matlab中的uigetfile函数输入服装图片,通过get函数实现颜色提取结果的展示,利用write函数保存服装颜色特征信息,给出系统退出代码。实验结果表明所设计系统能够直观展示服装颜色特征,且颜色特征提取结果可靠。     

基于特征提取算法的交互式英汉翻译系统设计

针对传统交互式英语翻译系统进行英语翻译过程中英语语义语境不明显,最优特征语义选定过程中没有达到最优翻译解,且翻译准确度较低的问题,设计基于特征提取算法的交互式英汉翻译系统。通过引入特征提取算法进行最优翻译解的选取,构建语义映射模型对英汉翻译进行交互最优翻译。仿真实验结果表明,设计的基于特征提取算法的交互式英汉翻译系统能够得到最优解翻译。     

基于MEEMD-HHT的分布式光纤振动传感系统信号特征提取方法

实际应用中,分布式光纤振动传感系统所测信号多为非平稳随机信号,对其进行模式识别的关键是准确获取信号的幅值-时间-频率瞬时特征。现有的相关研究表明,经验模态分解EMD方法结合希尔伯特变换可获得所测信号中固有模态分量的瞬时能量和瞬时频率,但存在模态混叠问题,后续改进的总体经验模态分解EEMD方法存在伪分量,重构误差大,互补经验模态分解CEEMD方法减小了重构误差的同时增加了运算量,无法保证特征提取与分类的效率与准确性。文中基于改进型经验模态分解方法结合希尔伯特变换MEEMD-HHT方法实现分布式光纤振动传感系统的特征提取,引入的排列熵的评价机制优化了分解过程中随机噪声迭代次数,通过仿真分析与实验对比,验证了该方法可有效解决上述方法中存在的问题,使系统在处理时间、特征准确度等性能皆有提高。实验结果表明,所提出的方法对于单频振动信号平均特征提取准确率达99.2%;对于混频振动信号平均特征提取准确率达98.1%,相对于EMD和CEEMD分别提高15.6%和7%,算法平均耗时最短,为3.8259 s,为分布式光纤振动传感系统的信号特征提取提供了一种可靠、高效的方法。     

On techniques for detecting circumscribed masses in mammograms

A method for detecting one type of breast tumor, circumscribed masses, in mammograms is presented. It relies on a combination of criteria used by experts, including the shape, brightness contrast, and uniform density of tumor areas. The method uses modified median filtering to enhance mammogram images and template matching to detect the tumors. In the template matching step, suspicious areas are identified by thresholding the cross-correlation values, and a percentile method is used to determine a threshold for each film. In addition, two tests are used to remove false alarms from the resulting candidates. The results obtained by applying these techniques to a set of test images are described. They are judged encouraging.     

基于支持向量机的超声乳腺肿瘤图像计算机辅助诊断系统

支持向量机与其他16种常见分类器相比具有优异的分类性能.本文提出一个基于支持向量机的超声乳腺肿瘤图像计算机辅助诊断系统.该系统提取了乳腺肿瘤图像归一化自相关系数、均方根系数、紧凑度等共26个纹理和几何特征.这些特征不但反映乳腺良性肿瘤和恶性肿瘤本质的区别,还考虑到医学超声图像为相干图像的特点.实验结果表明,该系统诊断精确度为93.03%,敏感性为94.30%,特殊性为91.59%,错误正比率92.80%,错误负比率为93.33%,接受特性曲线面积AUC为0.9669,性能优于较有影响的文献[1]提出的方案.     

Application of state variable techniques to optimal feature extraction

Optimal continuous linear feature extraction for the binary Gaussian pattern recognition or detection problem necessitates finding the double orthogonal expansion of the observable random process under hypothesis Hi, i = 1, 2. State variable techniques are utilized here to yield efficient computer-implementable procedures for obtaining the double orthogonal expansion.     

数字化工程中面向系统重用的信息提取

系统重用技术作为数字化工程的一部分,具有减少重复劳动的特点。为数字化工程实现缩短设计周期,降低开发成本和提高产品质量提供了重要支持。面向对象系统中的类模型、状态模型和交互模型在系统重用方面起着重要作用,为了提取被重用系统中的结构和功能信息,以类图、状态图和序列图为例,研究其数据在系统存储文件.sbs包中的组织结构,并且采用VC工具对一组实例图中的信息进行提取。结果表明,通过该存储文件获得的信息完全符合原设计意图,这为管理重用信息提供了数据支持。     

A computer-aided system for scaling topside ionograms

A computer-aided data-interpretation system is described that combines the abilities of a human operator for pattern recognition and decision making with those of a digital computer for rapid computation. It is designed for the specific task of converting ionograms to electron density N(h) profiles, but is sufficiently flexible in design and concept to be used, in principle, with any type of complex datum. The system consists of a small computer with an associated cathode-ray tube display, a human operator, and a large main computer. The small computer is used to digitize and store the ionogram tape and maintain the CRT display. The operator scales a trace representing one of the magnetoionic modes of the displayed ionogram, and commands the main computer to calculate the corresponding N(h) profile and the ionogram that would result from the computed profile. The operator then compares the remaining traces of the original and computed ionogram via the display, and either verifies the scaling or rescales the ionogram until the desired agreement has been achieved. The system permits the scaling, for N(h) analysis, of complex ionograms which otherwise could not be used for this purpose without great labor.     

An image segmentation based method for iris feature extraction

In this article, the local anomalistic blocks such as crypts, furrows, and so on in the iris are initially used directly as iris features. A novel image segmentation method based on intersecting cortical model （ICM） neural network was introduced to segment these anomalistic blocks. First, the normalized iris image was put into ICM neural network after enhancement. Second, the iris features were segmented out perfectly and were output in binary image type by the ICM neural network. Finally, the fourth output pulse image produced by ICM neural network was chosen as the iris code for the convenience of real time processing. To estimate the performance of the presented method, an iris recognition platform was produced and the Hamming Distance between two iris codes was computed to measure the dissimilarity between them. The experimental results in CASIA v1.0 and Bath iris image databases show that the proposed iris feature extraction algorithm has promising potential in iris recognition.