SVM for density estimation and application to medical image segmentation

A method of medical image segmentation based on support vector machine （SVM） for density estimation is presented. We used this estimator to construct a prior model of the image intensity and curvature profile of the structure from training images. When segmenting a novel image similar to the training images, the technique of narrow level set method is used. The higher dimensional surface evolution metric is defined by the prior model instead of by energy minimization function. This method offers several advantages. First, SVM for density estimation is consistent and its solution is sparse. Second, compared to the traditional level set methods, this method incorporates shape information on the object to be segmented into the segmentation process. Segmentation results are demonstrated on synthetic images, MR images and ultrasonic images.     

A general framework for progressive point-sampled geometry

INTRODUCTION With the recent advances in 3D digital photog- raphy and scanning technology, unstructured dense point sets sampled from the surfaces of physical ob- jects have become a new representation of geometric shapes. The shape thus specified is usually referred to as point-sampled geometry (Amenta and Kil, 2004; Liu et al., 2004; Pauly et al., 2003; Zwicker et al., 2002), in which two key techniques are involved: (1) The raw point data P is discrete in nature while the physical ob…     

基于一种新的曲率分析算法对散乱数据点云分块

给出了数据分块系统性方案, 即从仅含有三维坐标的散乱的点云中自动提取几何曲面特性. 首先基于局部基面参数化估算散乱数据点云的局部表面曲率分析是其方案的关键性技术. 再采用由高斯曲率和平均曲率的记号得到的8种曲面类型, 就形成初始数据分块. 通过区域增长法可以使粗略数据分块进一步被提取, 得到更小的噪声影响及更精确的区域划分. 其方案得到了实例验证, 具有较强的可操作性和实用性. 基于新曲率算法的分块方案使数据分块技术能够直接运用于散乱数据点云.     

肿瘤细胞膜免疫组化病理图像的自动分割

针对肿瘤细胞膜着色的免疫组化病理图像中组织灰度在RGB颜色通道中的分布形态特点,采用阈值分割方法将阳性产物、细胞、胞浆、间质及空隙等各组织区域分离;针对细胞核和阳性目标的灰度存在重叠现象,依据区域的面积、圆形度等几何形态特征建立二级分割模型,进一步将细胞核和阳性产物区分开来。在上述基础上,提出并实现一种肿瘤免疫组化病理图像的自动分割方法。结果表明,此方法速度快,分割效果理想。     

基于高斯分布估计的快速水平集分割方法研究

传统的CV模型只能用于分割灰度分布比较均匀、目标与背景灰度均值差异较大的图像,且因需要求解偏微分方程导致分割速度很慢.文章在传统快速水平集分割模型的基础上,将高斯分布估计引入速度项,使得快速水平集可以分割复杂的目标,并将HIS空间中的色调分量与强度分量进行融合,得到了一种彩色高斯快速模型.新算法具有分割速度快、可分割复杂目标的优点.     

Automatic tracing and segmentation of rat mammary fat pads in MRI image sequences based on cartoon-texture model

The growth patterns of mammary fat pads and glandular tissues inside the fat pads may be related with the risk factors of breast cancer.Quantitative measurements of this relationship are available after segmentation of mammary pads and glandular tissues.Rat fat pads may lose continuity along image sequences or adjoin similar intensity areas like epidermis and subcutaneous regions.A new approach for automatic tracing and segmentation of fat pads in magnetic resonance imaging (MRI) image sequences is presented,which does not require that the number of pads be constant or the spatial location of pads be adjacent among image slices.First,each image is decomposed into cartoon image and texture image based on cartoon-texture model.They will be used as smooth image and feature image for segmentation and for targeting pad seeds,respectively.Then,two-phase direct energy segmentation based on Chan-Vese active contour model is applied to partitioning the cartoon image into a set of regions,from which the pad boundary is traced iteratively from the pad seed.A tracing algorithm based on scanning order is proposed to accurately trace the pad boundary,which effectively removes the epidermis attached to the pad without any post processing as well as solves the problem of over-segmentation of some small holes inside the pad.The experimental results demonstrate the utility of this approach in accurate delineation of various numbers of mammary pads from several sets of MRI images.     

Automatic recognition of intersecting features of freeform sheet metal parts

This paper presents an approach for recognizing both isolated and intersecting geometric features of freeform surface models of parts, for the purpose of automating the process planning of sheet metal forming. The developed methodology has three major steps： subdivision of B-spline surfaces, detection of protrusions and depressions, and recognition of geometric features for sheet metal forming domain. The input geometry data format of the part is based on an IGES CAD surface model represented in the form of trimmed B-spline surfaces. Each surface is classified or subdivided into different curvature regions with the aid of curvature property surfaces obtained by using symbolic computation of B-spline surfaces. Those regions satisfying a particular geometry and topology relation are recognized as protrusion and depression （DP） shapes. The DP shapes are then classified into different geometric features using a rule-based approach. A verified feasibility study of the developed method is also presented.     

融合局部和全局信息的两相脑图分割模型

为了解决脑核磁共振图像中的灰度不均匀现象,提出了一个基于两相的脑图像分割模型．该模型在图像局部熵基础上融合局部高斯信息和全局信息的两相水平集来拟合图像信息,同时采用一种新的简单有效的初始化方法,进行脑灰质、脑白质、脑髓液分割．实验结果表明,该模型能有效地对脑灰质、脑白质、脑髓液进行分割,同时提供准确光滑的目标边界,解决了CV模型等对灰度不均匀图像分割失败的问题．     

一种快速自动多目标图像分割算法

针对进化多目标图像分割算法运行时间长且依赖人工挑选最优解的不足,提出一种快速自动多目标图像分割算法。首先使用自适应Mean-shift算法对图像进行预处理,将粗分割结果进行二次分割以提高运行速度;其次选择相互排斥的指标作为多目标的目标函数,并采用RM-MEDA框架对超像素颜色与纹理特征分别进行优化,同时对它们使用不同权值作为目标函数优化;最后由模糊模型从众多Pareto折中解集中自动选择满足实际分割要求的PS解。引入Mean-shift进行预分割,相对于标准的RM-MEDA,其运行速度提高近18%,由模糊模型推荐的Pareto解中,97%的情况符合分割要求。     

The relationship between approaches to teaching and approaches to studying: a two-level structural equation model for biology achievement in high school

Since the 1970s, a large body of research has reported on the differences between deep and surface approaches to student learning. More recently, however, this metaphor for students’ approaches to learning has been applied to the practice of teaching. Studies at the university level have identified two approaches to teaching: the information transmission/teacher-focused approach and the conceptual change/student-focused approach. The present study analyzes the relationship between teachers’ approaches to teaching and high school students’ approaches to learning. The data were analyzed by fitting a two-level structural equation model based on the hypothesis that student academic achievement is significantly determined by the way they study and that the way they study is partially determined by the way teachers teach. The participants were high school students (778 twelfth graders) enrolled in biology courses and their teachers (40 total). The same model was proposed at both levels (i.e., within and between levels) and fit the data quite well. As expected, within level, the effects of the ‘approaches to learning’ on ‘biology achievement’ regression were far larger than the corresponding effects at between level. The central findings suggest worthy directions for future research.     

Comparisons of two statistical approaches to study growth curves: The multilevel model and the latent curve analysis

This article compares two statistical approaches for modeling growth across time. The two statistical approaches are the multilevel model (MLM) and latent curve analysis (LCA), which have been proposed to depict change or growth adequately. These two approaches were compared in terms of the estimation of growth profiles represented by the parameters of initial status and the rate of growth. A longitudinal data set obtained from a school‐based substance‐use prevention trial for adolescents was used to illustrate the similarities and differences between the two approaches. The results indicated that the two approaches yielded very compatible results. The parameter estimates associated with regression weights are the same, whereas those associated with variances and covariances are similar. The MLM approach is easier for model specification and is more efficient computationally in yielding results. The LCA approach, however, has the advantage of providing model evaluation, that is, an overall test of goodness of fit, and is more flexible in modeling and hypothesis testing as demonstrated in this study.     

词位标注汉语分词技术详解

近年来基于字的词位标注的方法极大地提高了汉语分词的性能,该方法将汉语分词转化为字的词位标注问题,借助于优秀的序列标注模型,基于字的词位标注汉语分词方法逐渐成为分词的主要技术路线。本文简要介绍了词位标注汉语分词的基本思想,探析了基于条件随机场实现词位标注汉语分词的机理,并对采用四词位标注集,使用CRF＋＋0.53工具包实现字串序列词位标注进行了详解。最后在Bakeoff2006的评测语料上进行了封闭测试。     

Brain MRI segmentation using KFCM and Chan-Vese model

To extract region of interests (ROI) in brain magnetic resonance imaging (MRI) with more than two objects and improve the segmentation accuracy, a hybrid model of a kernel-based fuzzy c-means(KFCM) clustering algorithm and Chan-Vese (CV) model for brain MRI segmentation is proposed. The approach consists of two successive stages. Firstly, the KFCM is used to make a coarse segmentation, which achieves the automatic selection of initial contour. Then an improved CV model is utilized to subdivide the image. Fuzzy membership degree from KFCM clustering is incorporated into the fidelity term of the 2-phase piecewise constant CV model to obtain accurate multi-object segmentation. Experimental results show that the proposed model has advantages both in accuracy and in robustness to noise in comparison with fuzzy c-means (FCM) clustering, KFCM, and the hybrid model of FCM and CV on brain MRI segmentation.     

Low-cost sensorless control of four-switch, brushless DC motor drive with direct back-EMF detection

We propose a position sensorless control scheme for a four-switch, three-phase brushless DC motor drive, based on the zero crossing point detection of phase back-EMF voltages using newly defined error functions （EFs）. The commutation instants are 30° after detected zero crossing points of the EFs. Developed EFs have greater magnitude rather than phase or line voltages so that the sensorless control can work at a lower speed range. Moreover, EFs have smooth transitions around zero voltage level that reduces the commutation errors. EFs are derived from the filtered terminal voltages Vao and Vbo of two low-pass filters, which are used to eliminate high frequency noises for calculation of the average terminal voltages. The feasibility of the proposed sensorless control is demonstrated by simulation and experimental results.     

Similarity-based denoising of point-sampled surfaces

A non-local denoising （NLD） algorithm for point-sampled surfaces （PSSs） is presented based on similarities, including geometry intensity and features of sample points. By using the trilateral filtering operator, the differential signal of each sample point is determined and called ＂geometry intensity＂. Based on covariance analysis, a regular grid of geometry intensity of a sample point is constructed, and the geometry-intensity similarity of two points is measured according to their grids. Based on mean shift clustering, the PSSs are clustered in terms of the local geometry-features similarity. The smoothed geometry intensity, i.e., offset distance, of the sample point is estimated according to the two similarities. Using the resulting intensity, the noise component from PSSs is finally removed by adjusting the position of each sample point along its own normal direction. Experimental results demonstrate that the algorithm is robust and can produce a more accurate denoising result while having better feature preservation.     

Product information platform based on STEP

The intemational standard ISO 10303.called STEP.has been used to deal with problems in the ex change of product models and the associated data between difierent computer-aided systems.A platform based on STEP for managing product information is presented.This platform includes three components:a product geometry information model,a product feature model and a product visualization model.An information extracting pattern,in which information is extracted from low level elements to high level ones,is adopted in establishing the product geometry information model.Relative elements lists are created based on the extracted product information.With the traversing of these lists,feature extraction methods are proposed,which take advantage of boundary information in product model and avoid the determination of concavity and convexity of curves.Information correlating to features iS stored in a structure named as feature block and the product visualization model iS founded from it.The feature block is used in the platform for information communication and synchronous update among the three components.     

椭球面装饰铝板系统的设计

本文建立了椭球面上或偏离椭球面的经纬线长度、曲率半径、龙骨弯曲方向等几何参数的计算模型和方法。     

Finsler子流形的若干结果

研究Finsler子流形的若干性质.首先给出了Finsler流形能等距浸入到高维Minkowski空间中的一个新的简单的必要条件,即任何能等距浸入到Minkowski空间中的Finsler流形必定具有有限一致常数.其次以子流形的法曲率、T-曲率以及一致常数研究其象半径,当外围Finsler流形的旗曲率有上界时,得到了象半径估计,它是Riemannian几何中相关结果在Finsler几何中的推广.     

Research on cubic polynomial acceleration and deceleration control model for high speed NC machining

To satisfy the need of high speed NC （numerical control） machining, an acceleration and deceleration （acc/dec） control model is proposed, and the speed curve is also constructed by the cubic polynomial. The proposed control model provides continuity of acceleration, which avoids the intense vibration in high speed NC machining. Based on the discrete characteristic of the data sampling interpolation, the acc/dec control discrete mathematical model is also set up and the discrete expression of the theoretical deceleration length is obtained furthermore. Aiming at the question of hardly predetermining the deceleration point in acc/dec control before interpolation, the adaptive acc/dec control algorithm is deduced from the expressions of the theoretical deceleration length. The experimental result proves that the acc/dec control model has the characteristic of easy implementation, stable movement and low impact. The model has been applied in multi-axes high speed micro fabrication machining successfully.     

Intrusion detection using rough set classification

Recently machine learning-based intrusion detection approaches have been subjected to extensive researches because they can detect both misuse and anomaly. In this paper, rough set classification (RSC), a modern learning algorithm, is used to rank the features extracted for detecting intrusions and generate intrusion detection models. Feature ranking is a very critical step when building the model. RSC performs feature ranking before generating rules, and converts the feature ranking to minimal hitting set problem addressed by using genetic algorithm (GA). This is done in classical approaches using Support Vector Machine (SVM) by executing many iterations, each of which removes one useless feature. Compared with those methods, our method can avoid many iterations. In addition, a hybrid genetic algorithm is proposed to increase the convergence speed and decrease the training time of RSC. The models generated by RSC take the form of "IF-THEN" rules, which have the advantage of explication. Tests and compa