Content-based image retrieval with relevance feedback using random walks

In this paper, we propose a novel approach to content-based image retrieval with relevance feedback, which is based on the random walker algorithm introduced in the context of interactive image segmentation. The idea is to treat the relevant and non-relevant images labeled by the user at every feedback round as “seed” nodes for the random walker problem. The ranking score for each unlabeled image is computed as the probability that a random walker starting from that image will reach a relevant seed before encountering a non-relevant one. Our method is easy to implement, parameter-free and scales well to large datasets. Extensive experiments on different real datasets with several image similarity measures show the superiority of our method over different recent approaches.     

Pathological lung segmentation based on random forest combined with deep model and multi-scale superpixels

Accurate segmentation of lungs in pathological thoracic computed tomography (CT) scans plays an important role in pulmonary disease diagnosis. However, it is still a challenging task due to the variability of pathological lung appearances and shapes. In this paper, we proposed a novel segmentation algorithm based on random forest (RF), deep convolutional network, and multi-scale superpixels for segmenting pathological lungs from thoracic CT images accurately. A pathological thoracic CT image is first segmented based on multi-scale superpixels, and deep features, texture, and intensity features extracted from superpixels are taken as inputs of a group of RF classifiers. With the fusion of classification results of RFs by a fractional-order gray correlation approach, we capture an initial segmentation of pathological lungs. We finally utilize a divide-and-conquer strategy to deal with segmentation refinement combining contour correction of left lungs and region repairing of right lungs. Our algorithm is tested on a group of thoracic CT images affected with interstitial lung diseases. Experiments show that our algorithm can achieve a high segmentation accuracy with an average DSC of 96.45% and PPV of 95.07%. Compared with several existing lung segmentation methods, our algorithm exhibits a robust performance on pathological lung segmentation. Our algorithm can be employed reliably for lung field segmentation of pathologic thoracic CT images with a high accuracy, which is helpful to assist radiologists to detect the presence of pulmonary diseases and quantify its shape and size in regular clinical practices.

     

Random walks for image segmentation

A novel method is proposed for performing multilabel, interactive image segmentation. Given a small number of pixels with user-defined (or predefined) labels, one can analytically and quickly determine the probability that a random walker starting at each unlabeled pixel will first reach one of the prelabeled pixels. By assigning each pixel to the label for which the greatest probability is calculated, a high-quality image segmentation may be obtained. Theoretical properties of this algorithm are developed along with the corresponding connections to discrete potential theory and electrical circuits. This algorithm is formulated in discrete space (i.e., on a graph) using combinatorial analogues of standard operators and principles from continuous potential theory, allowing it to be applied in arbitrary dimension on arbitrary graphs     

改进FCM在交互式图像分割中的应用

为解决由于自然纹理的干扰而导致的分割图像边缘模糊问题,对模糊C均值聚类算法进行改进并应用于交互式图像分割中。用户通过输入种子点来获得目标和背景的主要特征,并将输入的种子点作为聚类中心点;提出全局空间相似性度量标准并引入Gabor能量滤波器来计算图像中各点到聚类中心的距离;算法首次引入边缘密度概念定义权重因子,根据图像特点,自适应地计算图像中任意一点的纹理特征和颜色特征在特征空间中所占比例,使得到的特征更加准确地描述图像的本质属性。对具有自然纹理背景的图像进行仿真实验,应用两种性能指标来比较本文所提算法与随机游走算法的分割精度。实验结果表明,本文算法分割精度高于模糊聚类和随机游走算法。     

基于稀疏表示和随机游走的磨玻璃型肺结节分割

肺结节是早期肺癌在影像学上的表现形式.磨玻璃型（Ground glass opacity,GGO）肺结节被认为是恶变可能性最大的一类结节之一.针对GGO结节边缘模糊、大小各异、形状不规则和灰度不均匀等造成分割准确率低问题,本文提出了一种基于稀疏表示和随机游走模型的分割算法.首先,利用测地距离和局部搜索策略,自动地选取了种子点.其次,联合8-!邻域和稀疏表示的K-!最近邻算法建立了新的图,避免了噪声的干扰.结合灰度、纹理、空间距离和稀疏系数构建了新的加权矩阵.最后,将标签限制项引入到随机游走的能量函数中.该算法分割准确性较高,鲁棒性较强.     

改进的交互式Otsu红外图像分割算法

为提高红外图像中目标分割的精度和抗噪性能,提出了一种改进的交互式Otsu图像分割算法。采用图像信息熵特征和类间方差特征对经典Otsu算法的阈值判别函数进行改进,获得的最优阈值能较好地将目标从背景中分割出来,且具有良好的边缘保持效果,提高了算法的分割精度。同时,针对红外图像目标单一的特点,采用交互式粗分割的思路,先在红外图像中提取包含目标的局部封闭区域,进而在提取的区域内进行改进的Otsu分割。通过对红外图像激光光斑目标提取过程的实验结果表明：改进的Otsu分割算法大大降低了背景噪声对分割算法的影响,提高了抗噪性能与分割精度,且最大程度地减少分割算法的运算量,并较好地保持了目标模糊边缘,分割效果优于传统的Otsu算法和相关的改进Otsu算法。     

基于K均值聚类的图割医学图像分割算法

图割是一种同时基于区域和边界的交互式图像分割算法。传统的基于高斯混合模型的图割具有时间慢和描述组织中灰度分布不准确的缺点。为此,提出一种基于K均值聚类的图割算法。通过用改进的图割来分割仿体的和真实的脑部核磁共振图像,显示出该方法的有效性。该方法不但能提高图割在分割时的速度,在有噪音和灰度不均匀的图像上也能在较短的时间内得到更准确且鲁棒的结果。     

Segmentation of piecewise stationary signals

Based on hierarchical clustering and dynamic programming, an algorithm for segmenting the piecewise stationary signal is developed. With some merits, ‘dispersion’ is proposed as a feature vector, which will be used both in the clustering method and dynamic programming. A test method is designed to evaluate the sensitivity of the distance measure constructed by the dispersion coefficients. The whole segmentation algorithm which constitutes three main stages is developed. In the first stage, called the feature extraction stage, the input signal is partitioned into several frames and the dispersion coefficients of each frame are evaluated. The second one is the hierarchical clustering stage, which clusters those frames based on the features—dispersion coefficients. The last stage is the dynamic segmentation stage, which finds the optimal change points by dynamic programming. Several simulations and real data were conducted and the results showed the satisfactory performance of this algorithm.     

图像分割的自适应交互核图割模型

为了克服图割模型算法在实现图像分割时需要人为选定参数,以及图割模型可能会陷入局部最小值的不足,考虑到交互图割是一种灵活的全局最优算法,提出了基于EM方法的交互核图割算法。数据映射到核空间,构造了新的目标函数,这样可以更有效地解决分类分割问题;为了估计交互图割所需要的参数以及图割算法所需要的各种阈值,采用EM算法来估计这些参数,避免人为随机选取可能造成的不利影响,因而该方法是一种自适应的分割算法。实验结果表明,相对于交互图割算法,该算法分割合成图像时具有更低的误分率,处理光学等图像时,分割结果更准确,保留图像细节信息的能力更强。     

Interactive characterization of granulated materials

This paper illustrates an interactive approach to digital characterization of granulated materials, based on mathematical morphology, digital topology and non-linear differential equations. The digital characterization is achieved using a three-step sequential process, namely: denoising, interactive segmentation and quantification, with emphasis in the interactive segmentation step, since it has significant impact on the characterization results. Denoising is used to extract foreground and background markers. Image segmentation results from an anisotropic region-growing and merging process based on immersion simulation. Region-growing evolves as a fully automated process, initiated at local image minima, while merging requires manual insertion of one single marker per region-of-interest. Non-significant regions are merged according to a hierarchy established by a set of markers and merging rules, which represent a robust solution to the over-segmentation problem inherent to immersion simulation algorithms. The sensitivity of the segmentation algorithm to local intensity variations allows the detection of the edges with high accuracy. Quantification is achieved using a suitable stereometric method, which computes the three-dimensional size distribution from the two-dimensional size distribution observed on the image. It is shown that this interactive approach represents a viable solution to digital characterization of granulated materials.     

一种基于局部歧义词网格的快速分词算法

提出了局部歧义词网格的概念,针对汉语分词中的覆盖歧义,提出了一种使用迭代算法训练覆盖歧义词典的算法,得到覆盖歧义候选词条词典。在此基础上提出了一种基于局部歧义词网格的、能够检测汉语分词过程中产生的组合歧义和覆盖歧义的分词算法,该算法仅考虑存在歧义的局部歧义词网格,并将对覆盖歧义的处理简化为查询覆盖歧义候选词典,因此,该算法的时间复杂度大幅下降。实验结果表明,该算法能够实现快速的汉语分词,且其分词正确率能够达到97%以上。     

基于GAC模型实现交互式图像分割的改进算法

提出了一种改进的交互式图像分割算法。采用全变分去噪模型对图像进行预处理,在去除噪声的同时更好地保护了边缘;提出了一种对梯度模值进行曲率加权的边缘检测方法,采用该方法获得图像的边缘点集;将边缘点集中曲率较大的边缘点作为候选边界点推荐给用户;用户通过主观判断,在候选边界点中选择合适的"初始边界点",算法便可采用GAC模型完成对目标的分割。实验结果表明,改进算法提高了交互式图像分割的自动化程度,有效地减少了交互过程中的人工参与量。     

分词中的歧义处理

歧义处理是影响分词系统切分精度的重要因素,是自动分词系统设计中的一个最困难也是最核心的问题。该文利用一种统计的方法来解决交集型歧义字段的切分。     

基于T-prim模型的肺气管分割算法

针对现有高阶支气管分割算法计算成本过大或分割精度不足等问题,提出一种基于T-prim模型的肺气管树分割算法。通过形态学灰度重建对CT图像进行初步处理,使用区域生长算法得到主支气管;从马尔可夫随机场的角度对分水岭算法分割框架进行优化,得到优化的分割框架;利用主气管骨架提取自动获得种子节点,算法迭代构造出T-prim模型,利用优化的分割框架得到完整的肺气管树。通过与两种EXACT09竞赛算法的对比实验证明了该算法在不依赖于种子点的人工选择,不需要训练集的条件下,能以极低的泄漏量获得更完整的分割结果。     

基于趋势转折点的时间序列分段线性表示

分段线性表示是时间序列降维的有效方法。在总结分析序列趋势变化特点的基础上,提出了一种基于趋势转折点的时间序列分段线性表示算法。首先定义了趋势转折点作为时间序列分段点的备选集,以点到区域的距离度量趋势转折点的重要性,再根据给定的阈值选择重要趋势转折点作为分段点,对时间序列进行分段线性表示。通过与其他6种方法进行实验比较,结果表明：所提方法在具有较好的拟合质量和适应能力以及对转折点明显的序列,都表现出较强的抗噪声干扰能力。     

曲面数控加工中面向NURBS刀具路径生成的刀位点分段算法

曲面数控加工中,NURBS刀具路径生成技术需要在大量有序刀位点中提取适合于NURBS刀具路径表示的刀位点段,刀位点的分段质量是决定NURBS刀具路径生成的前提.通过分析NURBS刀具路径的特点,对由刀位点表示的刀具路径之间的连接方式和边界点进行分类,提出通过层次聚类法将刀具路径进行分段的算法.在此基础上,通过判断连接点的类型来提取合适的刀位点段以进行NURBS刀具路径的生成.实例结果表明,该算法分段可靠、快捷,对不同曲线曲面轮廓刀具轨迹点进行分段的适应性强,分段结果可以满足NURBS刀具路径的生成.     

Segmenting handwritten signatures at their perceptually importantpoints

A new algorithm for segmenting continuous handwritten signatures sampled by a digitizer is described. The segmentation points are found using a two-step procedure. The principal step is to construct a function that weights the perceptual importance of every signature point according to its specific neighboring points. The second step points out the various local maxima of this function that correspond to where the signature should be segmented. The method is well illustrated and tested on a number of signatures that require different kinds of segmentation decisions     

基于多尺度的贝叶斯模型显著性检测

针对传统基于贝叶斯模型的显著性检测算法存在准确率不理想的问题,提出了一种基于多尺度的贝叶斯模型显著性检测算法。通过超像素分割算法(SLIC)将原图分割成不同尺度的超像素,根据超像素边界信息得到背景种子,进而通过距离计算和多尺度融合得到背景先验;对原图进行颜色增强,采用Harris算子对增强图进行检测角点求得凸包,融合不同尺度下的超像素得到凸包先验;融合背景先验和凸包先验得到最终先验;利用颜色直方图和凸包计算似然概率;将最终先验和似然概率通过贝叶斯模型计算显著图。在公开数据集MSRA1000、ECSSD上与多种传统算法进行准确率和召回率对比,该算法有更好的表现。     

汉语自动分词中的歧义处理

歧义处理是影响分词系统切分精度的重要因素，是自动分词系统设计中的一个核心问题。本文介绍了一种新的分词算法，利用汉语句内相邻字之间的互信息及t-信息差这两个统计量，解决汉语自动分词中的歧义字段的切分问题。试验结果表明，该方法可以有效地提高歧义处理的正确率。     

应用条件随机场进行汉语词法分析研究

中文分词是一个困难的、重要的被广泛研究的序列数据建模问题.以往应用条件随机场进行汉语分词时,将分词转化为对汉字的标注,造成了大量的冗余的候选切分,以至于在分词过程中大大降低了分词的速度.提出了使用词图作为基础的标记序列来完成汉语的词法分析,这样充分利用了现有的词典资源,在属性框架的选择时也可以方便地融合语言知识,并且长度歧视及状态歧视方面的影响也被减到最小.提出了应用条件随机场来构建统一的汉语词法分析.