Global gene expression in skeletal muscle from well-trained strength and endurance athletes

PURPOSE: We used gene microarray analysis to compare the global expression profile of genes involved in adaptation to training in skeletal muscle from chronically strength-trained (ST), endurance-trained (ET), and untrained control subjects (Con). METHODS: Resting skeletal muscle samples were obtained from the vastus lateralis of 20 subjects (Con n = 7, ET n = 7, ST n = 6; trained [TR] groups >8 yr specific training). Total RNA was extracted from tissue for two color microarray analysis and quantative (Q)-PCR. Trained subjects were characterized by performance measures of peak oxygen uptake V?O 2peak) on a cycle ergometer and maximal concentric and eccentric leg strength on an isokinetic dynamometer. RESULTS: Two hundred and sixty-three genes were differentially expressed in trained subjects (ET + ST) compared with Con (P < 0.05), whereas 21 genes were different between ST and ET (P < 0.05). These results were validated by reverse transcriptase polymerase chain reaction for six differentially regulated genes (EIFSJ, LDHB, LMO4, MDH1, SLC16A7, and UTRN. Manual cluster analyses revealed significant regulation of genes involved in muscle structure and development in TR subjects compared with Con (P < 0.05) and expression correlated with measures of performance (P < 0.05). ET had increased whereas ST had decreased expression of gene clusters related to mitochondrial/oxidative capacity (P ?‰Currency sign 0.05). These mitochondrial gene clusters correlated with V?O2peak (P < 0.05). V?O2peak also correlated with expression of gene clusters that regulate fat and carbohydrate oxidation (P < 0.05). CONCLUSION: We demonstrate that chronic training subtly coregulates numerous genes from important functional groups that may be part of the long-term adaptive process to adapt to repeated training stimuli.

A Study of feature extraction algorithms for optical flow tracking

Sparse optical flow algorithms, such as the Lucas-Kanade approach, provide more robustness to noise than dense optical flow algorithms and are the preferred approach in many scenarios. Sparse optical flow algorithms estimate the displacement for a selected number of pixels in the image. These pixels can be chosen randomly. However, pixels in regions with more variance between the neighbours will produce more reliable displacement estimates. The selected pixel locations should therefore be chosen wisely. In this study, the suitability of Harris corners, Shi-Tomasi's “Good features to track", SIFT and SURF interest point extractors, Canny edges, and random pixel selection for the purpose of frame-by-frame tracking using a pyramidical Lucas-Kanade algorithm is investigated. The evaluation considers the important factors of processing time, feature count, and feature trackability in indoor and outdoor scenarios using ground vehicles and unmanned aerial vehicles, and for the purpose of visual odometry estimation.

BILATERAL EDGE DETECTORS

We propose to employ bilateral filters to solve the problem of edge detection. The proposed methodology presents an efficient and noise robust method for detecting edges. Classical bilateral filters smooth images without distorting edges. In this paper, we modify the bilateral filter to perform edge detection, which is the opposite of bilateral smoothing. The Gaussian domain kernel of the bilateral filter is replaced with an edge detection mask, and Gaussian range kernel is replaced with an inverted Gaussian kernel. The modified range kernel serves to emphasize dissimilar regions. The resulting approach effectively adapts the detection mask according as the pixel intensity differences. The results of the proposed algorithm are compared with those of standard edge detection masks. Comparisons of the bilateral edge detector with Canny edge detection algorithm, both after non-maximal suppression, are also provided. The results of our technique are observed to be better and noise-robust than those offered by methods employing masks alone, and are also comparable to the results from Canny edge detector, outperforming it in certain cases.

Framework for morphometric classification of cells in imaging flow cytometry

Imaging flow cytometry is an emerging technology that combines the statistical power of flow cytometry with spatial and quantitative morphology of digital microscopy. It allows high-throughput imaging of cells with good spatial resolution, while they are in flow. This paper proposes a general framework for the processing/classification of cells imaged using imaging flow cytometer. Each cell is localized by finding an accurate cell contour. Then, features reflecting cell size, circularity and complexity are extracted for the classification using SVM. Unlike the conventional iterative, semi-automatic segmentation algorithms such as active contour, we propose a noniterative, fully automatic graph-based cell localization. In order to evaluate the performance of the proposed framework, we have successfully classified unstained label-free leukaemia cell-lines MOLT, K562 and HL60 from video streams captured using custom fabricated cost-effective microfluidics-based imaging flow cytometer. The proposed system is a significant development in the direction of building a cost-effective cell analysis platform that would facilitate affordable mass screening camps looking cellular morphology for disease diagnosis. Lay description In this article, we propose a novel framework for processing the raw data generated using microfluidics based imaging flow cytometers. Microfluidics microscopy or microfluidics based imaging flow cytometry (mIFC) is a recent microscopy paradigm, that combines the statistical power of flow cytometry with spatial and quantitative morphology of digital microscopy, which allows us imaging cells while they are in flow. In comparison to the conventional slide-based imaging systems, mIFC is a nascent technology enabling high throughput imaging of cells and is yet to take the form of a clinical diagnostic tool. The proposed framework process the raw data generated by the mIFC systems. The framework incorporates several steps: beginning from pre-processing of the raw video frames to enhance the contents of the cell, localising the cell by a novel, fully automatic, non-iterative graph based algorithm, extraction of different quantitative morphological parameters and subsequent classification of cells. In order to evaluate the performance of the proposed framework, we have successfully classified unstained label-free leukaemia cell-lines MOLT, K562 and HL60 from video streams captured using cost-effective microfluidics based imaging flow cytometer. The cell lines of HL60, K562 and MOLT were obtained from ATCC (American Type Culture Collection) and are separately cultured in the lab. Thus, each culture contains cells from its own category alone and thereby provides the ground truth. Each cell is localised by finding a closed cell contour by defining a directed, weighted graph from the Canny edge images of the cell such that the closed contour lies along the shortest weighted path surrounding the centroid of the cell from a starting point on a good curve segment to an immediate endpoint. Once the cell is localised, morphological features reflecting size, shape and complexity of the cells are extracted and used to develop a support vector machine based classification system. We could classify the cell-lines with good accuracy and the results were quite consistent across different cross validation experiments. We hope that imaging flow cytometers equipped with the proposed framework for image processing would enable cost-effective, automated and reliable disease screening in over-loaded facilities, which cannot afford to hire skilled personnel in large numbers. Such platforms would potentially facilitate screening camps in low income group countries; thereby transforming the current health care paradigms by enabling rapid, automated diagnosis for diseases like cancer.

Extração de linhas de imagens de sensoriamento remoto.

Essa dissertação tem o objetivo de verificar a contribuição de diferentes abordagens para extração de linhas, à classificação de imagens multiespectrais, com o possível uso na discriminação e mapeamento de classes de cobertura da terra. Nesse contexto, é efetuada a comparação entre diferentes técnicas de extração de características para extração de linhas de transmissão em áreas rurais, a saber, técnicas de realce utilizando variação de contraste e filtragem morfológica, bem como detecção de bordas utilizando filtro Canny e detector SUSAN, citando como técnica de extração de linhas a Transformada de Hough e Transformada de Radon, utilizando diferentes algoritmos, em imagens aéreas e de sensoriamento remoto. O processo de análise de imagens, com diferentes abordagens leva a resultados variados em diferentes tipos de coberturas do solo. Tais resultados foram avaliados e comparados produzindo tabelas de eficiência para cada procedimento. Estas tabelas direcionam a diferentes encaminhamentos, que vão variar de abordagem dependendo do objetivo final da extração das Linhas de Transmissão.

Desenvolvimento de ferramenta computacional para metrologia com microtomografia computadorizada

Esta tese apresentada uma proposta de desenvolvimento de uma ferramenta computacional para metrologia com microtomografia computadorizada que possa ser implantada em sistemas de microtomógrafos convencionais. O estudo concentra-se nas diferentes técnicas de detecção de borda utilizadas em processamento de imagens digitais.Para compreender a viabilidade do desenvolvimento da ferramenta optou-se por utilizar o Matlab 2010a. A ferramenta computacional proposta é capaz de medir objetos circulares e retangulares. As medidas podem ser horizontais ou circulares, podendo ser realizada várias medidas de uma mesma imagem, uma medida de várias imagens ou várias medidas de várias imagens. As técnicas processamento de imagens digitais implementadas são a limiarização global com escolha do threshold manualmente baseado no histograma da imagem ou automaticamente pelo método de Otsu, os filtros de passa-alta no domínio do espaço Sobel, Prewitt, Roberts, LoG e Canny e medida entre os picos mais externos da 1 e 2 derivada da imagem. Os resultados foram validados através de comparação com os resultados de teste realizados pelo Laboratório de Ensaios Mecânicos e Metrologia (LEMec) do Intstituto Politécnico do Rio de Janeiro (IPRJ), Universidade do Estado do Rio de Janeiro (UERJ), Nova Friburdo- RJ e pelo Serviço Nacional da Indústria Nova Friburgo (SENAI/NF). Os resultados obtidos pela ferramenta computacional foram equivalentes aos obtidos com os instrumentos de medição utilizados, demonstrando à viabilidade de utilização da ferramenta computacional a metrologia.

视频监控中若干关键问题的研究

随着网络带宽、计算机处理能力和存储容量的迅速提高，以及各种视频信息处理技术的出现，全程数字化、网络化的视频监控系统优势愈发明显。其高度的开放性、集成性和灵活性为视频监控系统和设备的整体性能提升创造了必要的条件；同时也为整个安防产业的发展提供了更加广阔的发展空间，崭新的应用模式和市场机遇不断涌现。视频监控系统过程向着大型、连续、综合化发展，形成了复杂监控过程，监测控制的要求越来越高，需要更高性能的系统和采用更优秀的控制手段，面临着不能用传统方法解决的新问题。本文概述了目前视频监控中面临的挑战，简要介绍了与视频监控相关的研究领域和研究现状，研究了视频监控中若干亟待解决的问题，主要取得了以下几个方面的研究成果： 第一，提出了基于AdaBoost的改进的人脸检测算法，针对AdaBoost算法的训练速度慢的问题，提出了基于阈值控制的训练方法；同时研究了AdaBoost算法人脸检测方法，利用肤色模型检测人脸区域，并对颜色模型进行了光照补偿。实验结果表明本文的算法具有较好的检测结果。 第二，提出了基于Canny算法的一般目标检测算法，提出了改进的Canny边缘检测算法，研究了Canny算法中噪声抑制的方法，采用改进中心加权的MTM算法有效的抑制噪声。针对Canny检测算法中阈值设置的问题，提出改进的Canny阈值补偿的方法。实验结果表明，改进的Canny算法相比原算法具有更好的目标检测性能。 第三，提出了一种基于均值漂移（Mean Shift）的改进的目标跟踪算法，通过搜索窗口带宽的计算，加权背景信息以及卡尔曼滤波器建模改进了跟踪算法，避免了均值漂移算法中的一些关键问题。对比实验结果表明，本文的改进方法相比原算法具有较好的性能。 第四，研究了视频监控中基于可扩展视频编码(SVC)的技术。首先讨论了视频监控中采用可扩展视频编码(SVC)的优势，探讨了视频监控中采用可扩展视频编码(SVC)的框架。然后针对于视频质量评估问题，设计并实现了基于可扩展视频编码(SVC)的视频质量评估系统Evalvid-SVC，研究了基于可扩展视频编码(SVC)的视频质量评估。 第五，研究并实现了视频数据安全传输技术。提出了基于Diameter的统一认证方案。任何用户想要获取视频资源，都必须通过AAA子系统的认证和授权，授予合法用户以特定的方式使用资源。另外，为了保证监控数据从监控前端安全地传送到视频监控客户端，本文提出了一种有效的保证视频数据安全传输的方案。设计的数据加密算法应用DES算法对前端设备采集到的音视频数据进行加密，并通过定时更新和RSA加密的方式保护和传输DES密钥。

单相机全向立体视觉研究

折反射全向成像系统是由普通透视相机和反射镜面组成的全向成像装置，可实时获取360°无需拼接的全景图像，近年来已成为研究热点并在视频会议、三维重建和移动机器人导航等领域有着广泛的应用。 本文主要对单相机全向立体视觉系统的设计、标定、匹配以及三维重建展开研究。介绍了一种可实时获取全向三维信息的折反射全向立体视觉光学装置OSVOD（Omnidirectional Stereo Vision Optical Device），OSVOD由两个双曲面镜和一个普通透视相机组成。其中两个双曲镜面上下同轴、间隔一定距离固定在一个玻璃筒内，下镜面中间开有一孔，上镜面通过下镜面的孔在相机像平面上成像，这样空间一点经上下反射镜的反射在像平面上有两个像点，用一个相机实现了立体视觉。两镜面的共同轴和相机镜头的光轴共线，共同焦点和镜头的光心重合，该配置能保证系统满足单一视点约束SVP（Single View-Point）。本结构配置也使系统的外极线呈一系列的放射线，对应点匹配简单。此外两镜面的间隔安装也使得系统的等效基线较长，从而具有较高的精度。 本文第一部分对当前的各种全向成像方法进行了简单介绍，并对各方法的特点做了归纳。第二部分介绍折反射全向视觉的研究现状，就各种反射镜面的成像特点做了对比。 第三部分介绍OSVOD的设计方法，包括机构的设计和镜面的设计，并对设计的结果做了误差分析。 第四部分是OSVOD的标定研究。给出了一种包括OSVOD中相机和镜面位置关系在内的系统参数的标定方法。该方法利用空间坐标已知的标定点在像平面上成的像，结合系统成像模型反算出标定点的空间坐标，再利用标定点的已知空间坐标和反算出的空间坐标建立方程，运用基于Levenberg-Marquardt的反向传播算法(backpropagation)标定相机与反射镜面间的安装偏差。该标定方法可推广到所有的折反射成像系统。 第五部分是基于全向图像的匹配研究。针对系统获取的立体图像对之间成像比例存在较大的差异，首先将图像展开成柱面投影图像，然后就下镜面成像展开的柱面做Canny边缘检测，得到了图像的边缘点；就得到的边缘点在展开的两幅柱面图像上做直接相关匹配。最后将获取的匹配点做一致性校验，并对一致性校验通过的匹配点做三维计算，生产稀疏的三维图像。 最后是结论和将来的工作展望。

基于VR的视景生成与基于小波变换的图像匹配算法研究

图像匹配是指把一个图像区域从另一个可能在不同时间，不同视点位置或者由不同传感器所获得的图像区域中确定出来或找到它们之间对应关系的一种重要的图像分析与处理技术。 实验结果验证是图像匹配中的一个重要环节。传统的图像匹配实验验证一般是在真实环境中进行。这种验证方式基本都面临实验周期长、成本高、安全性低、不易重复实验、可控性差、受气候条件、实验场地限制等诸多缺陷。在传统实物验证前，我们迫切需要寻找一种新的实验平台。虚拟现实技术的发展为我们构建新的实验平台提供了很好的技术支持代替传统方法。虚拟现实是一种由计算机和电子技术创造的看似真实的虚拟环境。通过多种传感设备，用户可根据自身的感觉，使用人的自然技能对虚拟世界中的物体进行考察和操作，参与其中的事件;同时提供视、听、触等直观而又自然的实时感知，并使参与者“沉浸”于虚拟环境中。作为虚拟现实行业的引导者，由Multigen-Paradigm公司开发的Creator三维建模软件和Vega实时可视化三维视景仿真软件已经广泛应用各种行业。本文采用这两种软件作为我们实验平台的开发工具。结果表明利用虚拟现实技术构建的仿真实验平台能够提供图像和数据，保证图像匹配实验的进行。 基于特征的图像匹配方法是提取图像中的一些特征点，然后借助这些特征点进行匹配。它能够较好的克服基于区域图像匹配算法易受实时性、光照、几何畸变等影响的缺点。边缘特征作为图像的一种基本特征已经广泛应用于图像匹配算法之中。传统的边缘检测方法(如Sobel、Prewitt和Canny算法)对噪声很敏感。 小波变换具有检测局部突变的能力，并且对图像噪声鲁棒，因此是检测图像边缘的有效工具。本文利用小波变换作为图像边缘检测工具，首先对图像分别做两方向的小波变换，进而得到二维小波变换的幅值和梯度，然后利用非极大值抑制方法检测二维小波变换的模极值点作为图像的边缘点。最后利用边缘直方图描述符实现模板边缘与实时图像边缘之间的匹配。 本文完成的主要工作如下： 1、 利用虚拟现实技术进行视景生成，构建图像匹配实验平台。输出实时图像以及加干扰的摄像机拍摄时所对应的水平角、俯仰角、自旋角、高度等指令参数。 2、 根据实时图像及对应参数，利用几何变换将前视图像转换为下视图像。 3、 在图像匹配中，我们在分析传统的边缘检测算子如Sobel、Prewitt、Canny等对噪声敏感的基础上，提出一种基于小波变换的图像边缘检测算法。 本文的实验环境为Vega、VC6.0++、Matlab7.0 实验结果表明本文所提出算法的有效性。

基于线性不变矩和角度向量的立体匹配算法

针对室内场景双目立体匹配有别于一般场景立体匹配的特殊性,提出了一种计算简便、准确度高的立体图像匹配算法。该算法首先利用canny算子检测物体的边缘,根据边缘的线性不变矩寻找出目标物体,然后提取出目标物体轮廓的特征点,利用角度直方图计算出左右图像的旋转角度,最后利用角度向量实现左右图像的对应像素点的匹配。线性不变矩有效地将计算复杂度由二维降低到一维,大大降低了计算量。角度向量的提出降低了特征点匹配的复杂度,而且计算简便,准确率高。实验表明,该算法对图像的缩放、旋转、平移均免疫,具有较高的识别精度和良好的抗干扰性,计算效率高于传统方法,有着较高的应用价值。

非结构环境下边缘检测参数自动优化方法

分析了基于梯度的边缘检测算法,针对其在参数确定自主能力不高的问题,提出一种新的基于大津法和统计理论的自适应边缘提取方法,通过对一组参数进行统计优化,自适应地确定边缘检测的全局最优参数,最后,采用Canny边缘检测算子对本算法进行验证.实验结果表明,本文提出的非结构环境下目标自适应边缘提取方法能够有效地抑制噪声,自适应地确定图像最优边缘参数,提高了边缘定位精度。

非结构环境下的自适应目标边缘提取方法

深入分析了经典的Canny边缘检测算法,针对其在参数确定的自主能力不高的问题,提出一种新的基于大津法和统计理论的自适应边缘提取方法,通过对一组参数进行了统计优化,自适应地确定边缘检测的全局最优参数。实验结果表明本文提出的非结构环境下目标自适应边缘提取方法能够有效地抑制噪声,自适应地确定最优边缘提取参数,提高了边缘定位精度。最后,通过实验表明,本文提出的方法在环境信息未知月球探测应用中具有较高边缘检测性能。

基于Zernike正交矩的图像亚像素边缘检测算法改进

介绍了Zernike矩及基于Zernike矩的图像亚像素边缘检测原理,针对Ghosal提出的基于Zernike矩的亚像素图像边缘检测算法检测出的图像存在边缘较粗及边缘亚像素定位精度低等不足,提出了一种改进算法.推导了7×7 Zernike矩模板系数,提出一种新的边缘判断依据.改进的算法能较好检测图像边缘并实现了较高的边缘定位.最后,设计了3组不同的实验.实验结果同Canny算子及Ghosal算法相比,证明了改进算法的优越性.

基于计算机视觉的轮毂自动识别与分类

随着全球汽车产量的逐年提高，相应地轮毂市场需求也在迅猛增长，国内80%的汽车轮毂是由铸造而成，铸造而成的轮毂需要进行精加工，而由混流生产线生产的轮毂在再加工时若对其类型进行正确划分将提高生产效率。以往的类型识别是依靠人工识别，工人劳动负荷较大，产生的视觉疲劳和轮毂生产线的速度等多方面原因致使误判情况时常发生，因此，轮毂企业急需一种自动的轮毂识别分类系统来提高生产效率。 计算机视觉是利用计算机对景物的图像进行识别，以实现对人视觉功能的扩展的一门高新技术。利用这一技术可以解决许多工业图像识别和检测环节的问题，以取代落后的人工识别，提高识别效率和工业自动化水平。 基于计算机视觉的轮毂自动识别与分类系统首先由图像采集系统从生产流水线上采集到要处理轮毂的图片，之后进行图像的预处理，包括轮毂图像区域的提取，图像去噪等操作，构造了四个具有平移不变性、比例不变性和幅度线性变换不变性的一维不变量，用以表征轮毂图像的灰度特征，利用边缘检测Canny算子提取轮毂外圆轮廓，之后对外圆轮廓运用改进的Hough变换方法计算出轮毂最大直径，最后运用马氏距离方法判别出轮毂类型。 本文对基于计算机视觉的轮毂自动识别与分类系统进行了总体设计，实现了图像采集、图像处理与轮毂型号识别等基本功能，克服了传统的人工识别的弊端，适应了混流生产线上快速识别分类的需要，开展了面向型号识别的计算机视觉系统的研究工作，为实现工业生产现代新型的图像识别技术做了一些有益的尝试和探讨。

Finding Edges and Lines in Images

The problem of detecting intensity changes in images is canonical in vision. Edge detection operators are typically designed to optimally estimate first or second derivative over some (usually small) support. Other criteria such as output signal to noise ratio or bandwidth have also been argued for. This thesis is an attempt to formulate a set of edge detection criteria that capture as directly as possible the desirable properties of an edge operator. Variational techniques are used to find a solution over the space of all linear shift invariant operators. The first criterion is that the detector have low probability of error i.e. failing to mark edges or falsely marking non-edges. The second is that the marked points should be as close as possible to the centre of the true edge. The third criterion is that there should be low probability of more than one response to a single edge. The technique is used to find optimal operators for step edges and for extended impulse profiles (ridges or valleys in two dimensions). The extension of the one dimensional operators to two dimentions is then discussed. The result is a set of operators of varying width, length and orientation. The problem of combining these outputs into a single description is discussed, and a set of heuristics for the integration are given.