利索-全球专利检索

  1. 登录
  2. 注册

搜索结果

46 条记录 (耗时 0.031 秒)

    Multi-level contextual learning of data
    2.
    发明授权
    Multi-level contextual learning of data 有权
    数据的多层次上下文学习

    公开(公告)号：US08724866B2

    公开(公告)日：2014-05-13

    申请号：US12962901

    申请日：2010-12-08

    申请人： Dijia Wu ,  Le Lu ,  Jinbo Bi ,  Yoshihisa Shinagawa ,  Marcos Salganicoff

    发明人： Dijia Wu ,  Le Lu ,  Jinbo Bi ,  Yoshihisa Shinagawa ,  Marcos Salganicoff

    IPC分类号： G06K9/00

    CPC分类号： G06K9/4638 ,  G06K2209/05

    摘要： Described herein is a framework for automatically classifying a structure in digital image data are described herein. In one implementation, a first set of features is extracted from digital image data, and used to learn a discriminative model. The discriminative model may be associated with at least one conditional probability of a class label given an image data observation Based on the conditional probability, at least one likelihood measure of the structure co-occurring with another structure in the same sub-volume of the digital image data is determined. A second set of features may then be extracted from the likelihood measure.

    摘要翻译： 这里描述了用于自动分类数字图像数据中的结构的框架。 在一个实现中，从数字图像数据中提取第一组特征，并用于学习辨别模型。 鉴别模型可以与给定图像数据观察的类标签的至少一个条件概率相关联。基于条件概率，与数字的相同子体积中的另一结构共同出现的结构的至少一个似然度量 确定图像数据。 然后可以从似然度量中提取第二组特征。

    Multi-Level Contextual Learning of Data
    3.
    发明申请
    Multi-Level Contextual Learning of Data 有权
    数据的多层次上下文学习

    公开(公告)号：US20110075920A1

    公开(公告)日：2011-03-31

    申请号：US12962901

    申请日：2010-12-08

    申请人： Dijia Wu ,  Le Lu ,  Jinbo Bi ,  Yoshihisa Shinagawa ,  Marcos Salganicoff

    发明人： Dijia Wu ,  Le Lu ,  Jinbo Bi ,  Yoshihisa Shinagawa ,  Marcos Salganicoff

    IPC分类号： G06K9/62

    CPC分类号： G06K9/4638 ,  G06K2209/05

    摘要： Described herein is a framework for automatically classifying a structure in digital image data are described herein. In one implementation, a first set of features is extracted from digital image data, and used to learn a discriminative model. The discriminative model may be associated with at least one conditional probability of a class label given an image data observation Based on the conditional probability, at least one likelihood measure of the structure co-occurring with another structure in the same sub-volume of the digital image data is determined. A second set of features may then be extracted from the likelihood measure.

    摘要翻译： 这里描述了用于自动分类数字图像数据中的结构的框架。 在一个实现中，从数字图像数据中提取第一组特征，并用于学习辨别模型。 鉴别模型可以与给定图像数据观察的类标签的至少一个条件概率相关联。基于条件概率，与数字的相同子体积中的另一结构共同出现的结构的至少一个似然度量 确定图像数据。 然后可以从似然度量中提取第二组特征。

    System and method for computer aided detection via asymmetric cascade of sparse linear classifiers
    4.
    发明申请
    System and method for computer aided detection via asymmetric cascade of sparse linear classifiers 有权
    通过稀疏线性分类器的不对称级联进行计算机辅助检测的系统和方法

    公开(公告)号：US20070110292A1

    公开(公告)日：2007-05-17

    申请号：US11592869

    申请日：2006-11-03

    申请人： Jinbo Bi ,  Senthil Periaswamy ,  Kazunori Okada ,  Toshiro Kubota ,  Glenn Fung ,  Marcos Salganicoff ,  R. Rao

    发明人： Jinbo Bi ,  Senthil Periaswamy ,  Kazunori Okada ,  Toshiro Kubota ,  Glenn Fung ,  Marcos Salganicoff ,  R. Rao

    IPC分类号： G06K9/00

    CPC分类号： G06K9/6256 ,  G06K9/6282

    摘要： A method for computer aided detection of anatomical abnormalities in medical images includes providing a plurality of abnormality candidates and features of said abnormality candidates, and classifying said abnormality candidates as true positives or false positives using a hierarchical cascade of linear classifiers of the form sign(wTx+b), wherein x is a feature vector, w is a weighting vector and b is a model parameter, wherein different weights are used to penalize false negatives and false positives, and wherein more complex features are used for each successive stage of said cascade of classifiers.

    摘要翻译： 一种用于计算机辅助检测医学图像中的解剖异常的方法，包括提供所述异常候选的多个异常候选和特征，并且使用形式符号（w）的线性分类器的分级级联将所述异常候选分类为真阳性或假阳性 x + b），其中x是特征向量，w是加权向量，b是模型参数，其中不同的权重用于惩罚假否定和假肯定，并且其中更复杂的特征 用于分级器级联的每个连续阶段。

    Method and apparatus for illuminating and imaging eyes through eyeglasses
    5.
    发明授权
    Method and apparatus for illuminating and imaging eyes through eyeglasses 失效
    用于通过眼镜照射和成像眼睛的方法和装置

    公开(公告)号：US6069967A

    公开(公告)日：2000-05-30

    申请号：US964359

    申请日：1997-11-04

    申请人： J. Michael Rozmus ,  Marcos Salganicoff

    发明人： J. Michael Rozmus ,  Marcos Salganicoff

    IPC分类号： A61B3/14 ,  G06K9/20 ,  G06K9/00

    CPC分类号： A61B3/14 ,  G06K9/2018

    摘要： A reliable method of illuminating and imaging an eye through eyeglasses uses a monochromatic light source with the smallest possible source area, a camera with an imager that exhibits minimal blooming, and a narrow-bandwidth optical bandpass filter to filter out most of the ambient illumination while passing most of the light from the system's own illuminator. In an alternative embodiment, a partially-transparent mirror is used to make the light source appear to be on the optical axis of the camera as viewed from the subject's eye.

    摘要翻译： 通过眼镜照射和成像眼睛的可靠方法使用具有尽可能最小的源区域的单色光源，具有最小亮度的成像器的相机和用于滤出大部分环境照明的窄带宽光带通滤波器， 通过系统自己的照明器的大部分光线。 在替代实施例中，使用部分透明的反射镜来使得从对象眼睛观察到的光源看起来在照相机的光轴上。

    Method of measuring the focus of close-up images of eyes
    6.
    发明授权
    Method of measuring the focus of close-up images of eyes 失效
    测量眼睛特写图像焦点的方法

    公开(公告)号：US5953440A

    公开(公告)日：1999-09-14

    申请号：US982364

    申请日：1997-12-02

    申请人： Guang Hua Zhang ,  Marcos Salganicoff

    发明人： Guang Hua Zhang ,  Marcos Salganicoff

    IPC分类号： A61B5/117 ,  G06K9/00 ,  G06K9/48 ,  G06K9/40

    CPC分类号： A61B5/117

    摘要： In a method of determining whether an image of an eye is in focus a set of pixels is selected along a line passing through the pupil/iris boundary such that the set contains at least 5 iris portion pixels and at least 5 pupil portion pixels. Statistical values, preferably median values, are computed for all iris pixels in the selected set and for all pupil pixels in the selected set. The step size between the iris pixels and the pupil pixels is computed and absolute gradient values are computed for each pixel. The pixel having a largest absolute gradient value is excluded and an average of the absolute gradient values of the remaining pixels is found. If that average divided by the step size is greater than 0.5 the image is in focus and can be used for identifying a subject whose eye is in the image using iris identification techniques.

    摘要翻译： 在确定眼睛的图像是否聚焦的方法中，沿着通过瞳孔/虹膜边界的线选择一组像素，使得该组包含至少5个虹膜部分像素和至少5个瞳孔部分像素。 对所选集合中的所有虹膜像素和所选集合中的所有瞳孔像素计算统计值，优选中值。 计算虹膜像素和瞳孔像素之间的步长，并为每个像素计算绝对梯度值。 排除具有最大绝对梯度值的像素，并且找到剩余像素的绝对梯度值的平均值。 如果该平均值除以步长大于0.5，则该图像被聚焦，并且可以用于使用虹膜识别技术识别眼睛在图像中的被摄体。

    Matching of regions of interest across multiple views
    8.
    发明授权
    Matching of regions of interest across multiple views 有权
    在多个视图中匹配感兴趣的区域

    公开(公告)号：US08885898B2

    公开(公告)日：2014-11-11

    申请号：US13267095

    申请日：2011-10-06

    申请人： Meizhu Liu ,  Le Lu ,  Vikas C. Raykar ,  Marcos Salganicoff ,  Matthias Wolf

    发明人： Meizhu Liu ,  Le Lu ,  Vikas C. Raykar ,  Marcos Salganicoff ,  Matthias Wolf

    IPC分类号： G06K9/00 ,  G06T7/00 ,  G06K9/62 ,  A61B6/03 ,  A61B8/13

    CPC分类号： G06K9/6215 ,  A61B6/032 ,  A61B6/037 ,  A61B8/13 ,  G06T7/0014 ,  G06T7/30 ,  G06T2207/10072 ,  G06T2207/20081 ,  G06T2207/30032

    摘要： Described herein is a framework for multi-view matching of regions of interest in images. According to one aspect, a processor receives first and second digitized images, as well as at least one CAD finding corresponding to a detected region of interest in the first image. The processor determines at least one candidate location in the second image that matches the CAD finding in the first image. The matching is performed based on local appearance features extracted for the CAD finding and the candidate location. In accordance with another aspect, the processor receives digitized training images representative of at least first and second views of one or more regions of interest. Feature selection is performed based on the training images to select a subset of relevant local appearance features to represent instances in the first and second views. A distance metric is then learned based on the subset of local appearance features. The distance metric may be used to perform matching of the regions of interest.

    摘要翻译： 这里描述了用于图像中感兴趣区域的多视图匹配的框架。 根据一个方面，处理器接收第一和第二数字化图像，以及对应于第一图像中检测到的感兴趣区域的至少一个CAD查找。 处理器确定与第一图像中的CAD查找匹配的第二图像中的至少一个候选位置。 基于为CAD查找和候选位置提取的本地外观特征执行匹配。 根据另一方面，处理器接收代表一个或多个感兴趣区域的至少第一和第二视图的数字化的训练图像。 基于训练图像执行特征选择，以选择相关局部外观特征的子集来表示第一和第二视图中的实例。 然后基于局部外观特征的子集来学习距离度量。 距离度量可以用于执行感兴趣区域的匹配。

    Reducing false positives for automatic computerized detection of objects
    9.
    发明授权
    Reducing false positives for automatic computerized detection of objects 有权
    减少自动计算机检测物体的误报

    公开(公告)号：US08160336B2

    公开(公告)日：2012-04-17

    申请号：US11516213

    申请日：2006-09-06

    申请人： Jianming Liang ,  Zhiyun Xue ,  Marcos Salganicoff ,  Matthias Wolf

    发明人： Jianming Liang ,  Zhiyun Xue ,  Marcos Salganicoff ,  Matthias Wolf

    IPC分类号： G06K9/00

    CPC分类号： G06T7/0012

    摘要： A computer-implemented method for identifying an object of interest includes providing input data including an image and a candidate for the object of interest in the image, extracting a boundary of the candidate, and extracting a segment of a region of interest containing the candidate. The method further includes determining a plurality of features of an extracted segment of the region of interest containing the candidate, and outputting the object of interest, wherein the object of interest is characterized by the plurality of features, wherein the object of interest and the plurality of features are stored as computer-readable code.

    摘要翻译： 用于识别感兴趣对象的计算机实现的方法包括提供包括图像中的图像和感兴趣对象的候选者的输入数据，提取候选者的边界，以及提取包含候选的感兴趣区域的片段。 该方法还包括确定包含该候选者的所述感兴趣区域的提取段的多个特征，并输出所述感兴趣对象，其中所述感兴趣对象由所述多个特征表征，其中所述感兴趣对象和所述多个特征 功能存储为计算机可读代码。

    Matching of Regions of Interest Across Multiple Views
    10.
    发明申请
    Matching of Regions of Interest Across Multiple Views 有权
    多个意见区域的匹配

    公开(公告)号：US20120088981A1

    公开(公告)日：2012-04-12

    申请号：US13267095

    申请日：2011-10-06

    申请人： Meizhu Liu ,  Le Lu ,  Vikas C. Raykar ,  Marcos Salganicoff ,  Matthias Wolf

    发明人： Meizhu Liu ,  Le Lu ,  Vikas C. Raykar ,  Marcos Salganicoff ,  Matthias Wolf

    IPC分类号： A61B5/00 ,  G06K9/00

    CPC分类号： G06K9/6215 ,  A61B6/032 ,  A61B6/037 ,  A61B8/13 ,  G06T7/0014 ,  G06T7/30 ,  G06T2207/10072 ,  G06T2207/20081 ,  G06T2207/30032

    摘要： Described herein is a framework for multi-view matching of regions of interest in images. According to one aspect, a processor receives first and second digitized images, as well as at least one CAD finding corresponding to a detected region of interest in the first image. The processor determines at least one candidate location in the second image that matches the CAD finding in the first image. The matching is performed based on local appearance features extracted for the CAD finding and the candidate location. In accordance with another aspect, the processor receives digitized training images representative of at least first and second views of one or more regions of interest. Feature selection is performed based on the training images to select a subset of relevant local appearance features to represent instances in the first and second views. A distance metric is then learned based on the subset of local appearance features. The distance metric may be used to perform matching of the regions of interest.

    摘要翻译： 这里描述了用于图像中感兴趣区域的多视图匹配的框架。 根据一个方面，处理器接收第一和第二数字化图像，以及对应于第一图像中检测到的感兴趣区域的至少一个CAD查找。 处理器确定与第一图像中的CAD查找匹配的第二图像中的至少一个候选位置。 基于为CAD查找和候选位置提取的本地外观特征执行匹配。 根据另一方面，处理器接收代表一个或多个感兴趣区域的至少第一和第二视图的数字化的训练图像。 基于训练图像执行特征选择，以选择相关局部外观特征的子集来表示第一和第二视图中的实例。 然后基于局部外观特征的子集来学习距离度量。 距离度量可以用于执行感兴趣区域的匹配。