基于连通区域的传真图像版面分割与分类算法

针对传真图像,为了提高版面分割与分类准确率,提高处理速度,以连通区域为处理元素,通过合理设定阈值,将水平和垂直相邻连通区域进行合并,快速准确地分割图像。并且将版面分割过程与分类过程相结合,根据连通区域的大小建立矩阵,提取能够表征区域信息的八维特征,然后使用BP神经网络将版面区域分为文字区域和非文字区域两类。实验中得到版面分割准确率为89.2%,版面分类准确率为94.22%。实验结果证明,该算法能够快速准确地对传真图像进行分割和分类,具有较强的实用价值。     

连通区的页面分割与分类方法

页面分割与分类是文档处理的关键步骤，但目前多数方法对页面的块和倾斜进行了限制，文中提出一种新的基于连通区的页面分割与分类方法，首行采用快速算法抽取页面内的连通区，然后利用改进的PLSA算法分割页面，并根据连通区的分布情况以及块的特征对块进行分类，该方法页面分割与分类紧密结合，充分考虑到块的局部特征，保证块分类的正确性，大大提高了算法效率。     

基于决策树的注采连通关系判别研究

判别注采连通关系的传统方法是由人工通过对比测井曲线, 逐个层位地进行连通关系的识别和分类, 这种方式工作量大、耗费劳动力多、判别效率较低。针对大庆油田注采连通关系判别的这种现状, 从油藏开发数据中提取并构建出注采井间的相对特征, 使用CART（分类回归树）算法建立了注采连通关系自动判别的决策树模型。实验结果表明, 该方法具有操作简单、判别速度快等优点, 在提高连通关系判别效率的同时保证了有较高的精度。     

海量遥感分类图连通域标记方法

针对遥感影像数据量大的特点,设计并实现了一种针对海量遥感影像分类图的连通域标记算法。通过对影像进行合理高效的分块,解决海量数据处理过程中对大内存的要求。对读入内存中的数据,采用了基于队列的种子填充算法进行连通域标记,通过采用贯序连通域标记算法中的冲突表机制,解决因分块之间隔离而造成的同一连通域标记不同的问题。该方法具有可以处理海量数据、对分类结果数据仅仅访问一次、无须对生成的结果进行重新标记就可以完成连通域的标记、可以获取连通域的基本统计信息等特点。实验结果证明了该方法的高效性。     

无线传感器网络连通恢复综述

针对无线传感器网络中连通恢复问题,分析和总结了近年来相关的主要方向和研究成果,同时对无线传感器网络连通恢复解决方法进行了分类和总结,并指出了其中的不足与未来的研究方向。     

基于Hopfield神经网络的输油站场操作算法

提出利用连通图来描述输油站场工艺流程。通过对连通图节点进行分类,将计算具体调度任务的操作方案问题转化为计算连通图的辐射状子图问题。在连通图基础上提出最优操作方案的数学模型,并提出利用Hopfield神经网络对数学模型进行求解。给出了神经网络的能量函数和对应的求解方法,实际计算表明,所提方法可行有效,能够减轻现场劳动强度,提高生产效率。     

基于扫描线的连通域搜索算法

基于扫描线的连通域搜索法是利用扫描线上相邻元素必然是连通域，以及扫描过程中相邻扫描线之间的关系的特点．给出了快速准确的实现连通域搜索的算法。本文给出算法描述。并讨论使用C＋＋实现中的困难，最后给出关键代码部分．即相关算法复杂度的比较。     

一种基于极大连通子图的相关度属性选择算法

本文提出一种基于极大连通子图的相关度属性选择算法,该算法利用极大连通子图求解方法得到最优属性子集,该属性子集满足相关度要求并且尽可能多地保留数据集中包含的信息。论文结合选定的分类算法验证该属性选择算法的应用效果,实验表明,该属性选择算法在分类算法的数据预处理过程中可明显提高分类准确率,与利用贪婪搜索策略的基于相关度的属性选择算法相比,应用该算法的分类算法准确率更高。     

基于数学形态学的带障碍约束的空间聚类算法研究

该文提出了一种基于数学形态学的带障碍约束的聚类算法。通过数学形态学的膨胀运算,进行连通区域的寻找,同时借助于进行膨胀运算的结构元素,确定障碍物与连通区域是否相交。算法与DBCluC算法不同的是:通过结构元素,大大减少了需要进行相交判断的点的数量,具有较高的时间效率。     

面向异构数据库集成的混合式元素分类技术研究

提出一种支持异构数据库集成的基于模式的元素分类技术,融合了基于语言和基于约束的分类方法,首先使用最大优先策略组合基于语言的分类算法形成复合元素名称分类器,然后将名称分类结果和元素模式约束转化为向量输入神经网络分类器,从而完成数据库元素分类.通过组合使用复合名称分类器和神经网络分类器,使得此方法可以应用于更复杂的分类环境.     

Segmentation and classification of mixed text/graphics/image documents

In this paper, a feature-based document analysis system is presented which utilizes domain knowledge to segment and classify mixed text/graphics/image documents. In our approach, we first perform a run-length smearing operation followed by the stripe merging procedure to segment the blocks embedded in a document. The classification task is then performed based on the domain knowledge induced from the primitives associated with each type of medium. Proper use of domain knowledge is proved to be effective in accelerating the segmentation speed and decreasing the classification error. The experimental study reveals the feasibility of the new technique in segmenting and classifying mixed text/graphics/image documents.     

一种基于HVS的DCT域图像自适应数字水印新算法

本文提出了一种基于人类视觉系统（HVS）的DCT域图像自适应数字水印算法，该算法可对原始宿主图像进行分块DC了变换，使用Robert算子对每一小块追行边缘检测提取，采用确定数值作为阀值对图像块进行分类，从而对嵌入的水印强度进行调整。添加水印时，对二值水印图像进行Arnold置乱、降维预处理，并将此序列嵌入到每块的DC了低频系数上。仿真实验结果表明，本文实现的水印具有不可见性，且对于常见的JPEG压缩、噪声、裁剪等攻击具有较好的鲁棒性。     

基于颜色分量间相关性的图像拼接篡改检测方法

由于目前数码相机在获取自然图像时,都存在着某种颜色滤波阵列（CFA）插值效应,使得图像颜色分量间具有很大的相关性。针对此问题提出基于CFA插值产生插值特性的图像拼接篡改检测方法。首先对图像颜色分量进行CFA插值预测,得到预测误差;然后计算图像块预测误差的局部加权方差得到图像块的CFA特征;最后根据高斯混合参数模型,对提取特征进行分类得到篡改区域。在标准拼接篡改图像数据集中的实验结果显示,此方法能够有效地检测出图像篡改区域的精确位置。     

基于四元数傅里叶梅林变换的旋转不变彩色纹理分类

提出一种基于四元数傅里叶梅林变换(Quaternion Fourier-Mellin Transform,QFMT)的旋转不变彩色纹理分类方法。该方法首先对彩色图像各分量图像进行对数极坐标变换,然后将经过变换后的3幅分量图像表示成四元数,并对其进行四元数傅里叶变换(Quaternion Fourier Transform,QFT),最后对幅度谱分别统计其环形特征量和楔形特征量作为纹理分类的特征向量,利用最近邻分类器进行分类。实验结果表明,本文提出的方法分类准确率更高,且具有良好的旋转不变纹理分析性能。     

Digital image splicing detection based on approximate run length

Image splicing is very common and fundamental in image tampering, which severely threatens the integrity and authenticity of images. As a result, the detection of image splicing is of great importance. In this paper, an approximate run length based scheme is proposed to detect this specific artifact. Firstly, the edge gradient matrix of an image is computed, and approximate run length is calculated along the edge gradient direction. Then, some features are constructed from the histogram of the approximate run length. To further improve the detection accuracy, the approximate run length is applied on the predict-error image and the reconstructed images based on DWT to obtain more features. Finally, support vector machine (SVM) is exploited to classify the authentic and spliced images using the constructed features. The experiment results demonstrate that the proposed approach can achieve a relatively high accuracy with less computational cost and fewer features when compared with other methods.     

基于灰度共生矩阵的断口图像识别

研究断口图像的识别是进行失效分析的一个重要组成部分,但普遍存在识别率不高的问题,为了有效提高识别率,提出一种基于灰度共生矩阵的断口识别方法,方法采用灰度共生矩阵法对金属断口图像进行特征提取,利用可分性判据来确定能影响纹理特征参数的构造因子(生成步长d,图像灰度级g和生成方向θ),并通过相关性分析来排除在表达金属断口图像纹理中冗余的特征参数,最后采用BP神经网络分类器进行分类,并进行仿真。结果证明,根据上述方法,对金属断口图像的分类是可行的,并且识别率也达到了满意的要求。     

基于连通域特征的维吾尔手写文本行分割

针对维吾尔文手写体文本中行分割问题,基于连通域大小将图像中文字分为三类,提出了自适应涂抹细化算法,对主体文本行进行定位;并对第三类连通域中相邻两文本行间粘连的字符进行切割;此外,利用重心范围内的邻域搜索算法,解决了剩余笔画的文本行归附问题。实验结果表明,该方法与常见的水平投影法,分段投影法,及涂抹方法相比具有更好的分割效果。     

Multi-look versus single-look processing of synthetic aperture radar images with respect to ocean wave spectra estimation

Abstract

Full-bandwidth C-band synthetic aperture radar (SAR) data are compared with 7-look and 3-look data. The peak-to-background ratio of the image intensity power spectrum describing the wave detectability is found to be on average 8-9dB higher for the 7-look data and 2-5dB higher for the 3-look data than the single-look data. This is mainly due to the decrease in the speckle noise level when going from single-look to multi-look processing. In addition, look-sum processing is evaluated against spectral-sum processing for various temporal look separations. A significant improvement in image spectral peak contrast is observed for the spectral-sum data versus the look-sum data, with increasing temporal separations between the looks. No such improvement is observed in the corresponding image spectral noise contrast parameter. These observations are in agreement with the spatial misregistration inherent in look-sum data. Finally, the acceleration contribution to the observed aximuth smearing in the spectra is found to be negligible compared with the velocity smearing contribution.     

A fuzzy topological framework for classifying image databases

The problem considered in this paper is how to classify image databases in terms of semantically coherent image categories. An image category (image concept) is represented by a set of images with visual and semantic similarities. We propose a topological framework to model each image concept and also classify images. Classification part utilizes a new form of fuzzy interior. To cope with uncertainties associated with feature extraction part and to achieve high classification acuity, the proposed fuzzy interior method takes into account the inclusion degree. This is accomplished by considering a variation of fuzzy subsethood used in the definition of fuzzy interior and by viewing image feature histograms as fuzzy sets. Because each image category is modeled independently, adding new categories to the system is both efficient and easily accomplished. The main contribution of this paper is the introduction of a fuzzy topological framework to classify image databases. © 2011 Wiley Periodicals, Inc.     

A differential-processing extraction approach to text and image segmentation

To efficiently store the information found in paper documents, text and non-text regions need to be separated. Non-text regions include half-tone photographs and line diagrams. The text regions can be converted (via an optical character reader) to a computer-searchable form, and the non-text regions can be extracted and preserved in compressed form using image-compression algorithms. In this paper, an effective system for automatically segmenting a document image into regions of text and non-text is proposed. The system first performs an adaptive thresholding to obtain a binarized image. Subsequently the binarized image is smeared using a run-length differential algorithm. The smeared image is then subjected to a text characteristic filter to remove error smearing of non-text regions. Next, baseline cumulative blocking is used to rectangularize the smeared region. Finally, a text block growing algorithm is used to block out a text sentence. The recognition of text is carried out on a text sentence basis.