利索-全球专利检索

  1. 登录
  2. 注册

搜索结果

215 条记录 (耗时 0.028 秒)

    Method and apparatus for tracking multiple objects in a video sequence
    1.
    发明授权
    Method and apparatus for tracking multiple objects in a video sequence 有权
    用于跟踪视频序列中的多个对象的方法和装置

    公开(公告)号：US06879705B1

    公开(公告)日：2005-04-12

    申请号：US09615971

    申请日：2000-07-14

    申请人： Hai Tao ,  Rakesh Kumar ,  Harpreet Singh Sawhney

    发明人： Hai Tao ,  Rakesh Kumar ,  Harpreet Singh Sawhney

    IPC分类号： G06K9/00 ,  G06T7/20

    CPC分类号： G06T7/20

    摘要： A method and apparatus for tracking multiple objects in a video sequence. The method defines a group of objects as a configuration, selects a configuration for a current video frame, predicts a configuration using a two-level process and computes the likelihood of the configuration. Using this method in an iterative manner on a sequence of frames, tracks the object group through the sequence.

    摘要翻译： 一种用于跟踪视频序列中的多个对象的方法和装置。 该方法将一组对象定义为配置，为当前视频帧选择配置，使用两级进程预测配置，并计算配置的可能性。 在一系列帧上以迭代的方式使用此方法，通过序列跟踪对象组。

    Dynamic depth recovery from multiple synchronized video streams
    3.
    发明授权
    Dynamic depth recovery from multiple synchronized video streams 有权
    从多个同步视频流动态深度恢复

    公开(公告)号：US06847728B2

    公开(公告)日：2005-01-25

    申请号：US10315291

    申请日：2002-12-09

    申请人： Hai Tao ,  Harpreet Singh Sawhney ,  Rakesh Kumar

    发明人： Hai Tao ,  Harpreet Singh Sawhney ,  Rakesh Kumar

    IPC分类号： G06T7/00 ,  G06K9/00

    CPC分类号： G06T7/579

    摘要： A method of generating a dynamic depth map for a sequence of images from multiple cameras models a scene as a collection of 3D piecewise planar surface patches induced by color based image segmentation. This representation is continuously estimated using an incremental formulation in which the 3D geometric, motion, and global visibility constraints are enforced over space and time. The proposed algorithm optimizes a cost function that incorporates the spatial color consistency constraint and a smooth scene motion model.

    摘要翻译： 为来自多个摄像机的图像序列生成动态深度图的方法将场景模拟为基于颜色的图像分割引起的3D分段平面表面贴片的集合。 使用增量公式连续估计这种表示，其中3D几何，运动和全局可见性约束在空间和时间上被强制执行。 所提出的算法优化了包含空间颜色一致性约束和平滑场景运动模型的成本函数。

    Method and apparatus for tracking moving objects in a sequence of two-dimensional images using a dynamic layered representation
    4.
    发明授权
    Method and apparatus for tracking moving objects in a sequence of two-dimensional images using a dynamic layered representation 有权
    使用动态分层表示来跟踪二维图像序列中的移动物体的方法和装置

    公开(公告)号：US06826292B1

    公开(公告)日：2004-11-30

    申请号：US09603666

    申请日：2000-06-23

    申请人： Hai Tao ,  Rakesh Kumar ,  Harpreet Singh Sawhney

    发明人： Hai Tao ,  Rakesh Kumar ,  Harpreet Singh Sawhney

    IPC分类号： G06K300

    CPC分类号： G06K9/32 ,  G06T7/215 ,  G06T7/277 ,  G06T2207/10016 ,  G06T2207/10032 ,  G06T2207/20116 ,  G06T2207/30236 ,  G06T2207/30252

    摘要： A system that tracks one or more moving objects in a sequence of video images employs a dynamic layer representation to represent the objects that are being tracked. The system concurrently estimates three components of the dynamic layer representation—layer segmentation, motion, and appearance—over time in a maximum a posteriori (MAP) framework. In order to enforce a global shape constraint and to maintain the layer segmentation over time, the subject invention imposes a prior constraint on parametric segmentation. In addition, the system uses a generalized Expectation-Maximization (EM) algorithm to compute an optimal solution. The system uses an object state that consists of representations of motion, appearance and ownership masks. The system applies a constant appearance model across multiple images in the video stream.

    摘要翻译： 跟踪视频图像序列中的一个或多个移动对象的系统采用动态层表示来表示正被跟踪的对象。 该系统在最大后验（MAP）框架中同时估计动态层表示层分段，运动和随时间出现的三个组成部分。 为了实施全局形状约束并且随着时间维持层分割，本发明对参数分割施加了先前的约束。 此外，该系统使用广义期望最大化（EM）算法来计算最优解。 系统使用由运动，外观和所有权掩码的表示组成的对象状态。 系统在视频流中跨多个图像应用恒定外观模型。

    System and method for detection of multi-view/multi-pose objects
    5.
    发明申请
    System and method for detection of multi-view/multi-pose objects 有权
    用于检测多视点/多姿态对象的系统和方法

    公开(公告)号：US20120002869A1

    公开(公告)日：2012-01-05

    申请号：US13134885

    申请日：2011-06-20

    申请人： Feng Han ,  Ying Shan ,  Harpreet Singh Sawhney ,  Rakesh Kumar

    发明人： Feng Han ,  Ying Shan ,  Harpreet Singh Sawhney ,  Rakesh Kumar

    IPC分类号： G06K9/62

    CPC分类号： G06K9/6256

    摘要： The present invention provides a computer implemented process for detecting multi-view multi-pose objects. The process comprises training of a classifier for each intra-class exemplar, training of a strong classifier and combining the individual exemplar-based classifiers with a single objective function. This function is optimized using the two nested AdaBoost loops. The first loop is the outer loop that selects discriminative candidate exemplars. The second loop, the inner loop selects the discriminative candidate features on the selected exemplars to compute all weak classifiers for a specific position such as a view/pose. Then all the computed weak classifiers are automatically combined into a final classifier (strong classifier) which is the object to be detected.

    摘要翻译： 本发明提供了一种用于检测多视点多姿态对象的计算机实现过程。 该过程包括针对每个类内样本的分类器的训练，强分类器的训练和将单个基于样本的分类器与单个目标函数组合。 使用两个嵌套的AdaBoost循环来优化此功能。 第一个循环是选择区分候选样本的外循环。 第二个循环，内循环选择所选样本上的鉴别候选特征，以计算特定位置（例如视图/姿态）的所有弱分类器。 然后将所有计算的弱分类器自动组合成最终分类器（强分类器），该分类器是要检测的对象。

    Method and apparatus for multi-view three dimensional estimation
    7.
    发明授权
    Method and apparatus for multi-view three dimensional estimation 有权
    用于多视图三维估计的方法和装置

    公开(公告)号：US06571024B1

    公开(公告)日：2003-05-27

    申请号：US09336319

    申请日：1999-06-18

    申请人： Harpreet Singh Sawhney ,  Rakesh Kumar ,  Yanlin Guo ,  Jane Asmuth ,  Keith James Hanna

    发明人： Harpreet Singh Sawhney ,  Rakesh Kumar ,  Yanlin Guo ,  Jane Asmuth ,  Keith James Hanna

    IPC分类号： G06K932

    CPC分类号： G06T7/248 ,  G06T7/246 ,  G06T2207/10016 ,  G06T2207/30244

    摘要： An apparatus and method for generating automated multi-view three dimensional pose and geometry estimation for the insertion of realistic and authentic views of synthetic objects into a real scene. A multi-view three dimensional estimation routine comprising the steps of feature tracking, pairwise camera pose estimation, computing camera pose for overlapping sequences and performing a global block adjustment to provide camera pose and scene geometric information for each frame of a scene. A match move routine may be used to insert a synthetic object into one frame of a video sequence based on the pose and geometric information of the frame, and calculate all other required object views of the synthetic object for the remaining frames using the pose and geometric information acquired as a result of the multi-view three dimensional estimation routine.

    摘要翻译： 一种用于产生用于将合成对象的真实和真实视图插入到真实场景中的自动多视图三维姿态和几何估计的装置和方法。 一种多视图三维估计程序，包括特征跟踪，配对摄像机姿态估计，重叠序列的计算摄像机姿态，以及执行全局块调整以为场景的每个帧提供摄像机姿态和场景几何信息的步骤。 可以使用匹配移动例程来基于帧的姿态和几何信息将合成对象插入到视频序列的一帧中，并且使用姿态和几何来计算剩余帧的合成对象的所有其他所需对象视图 作为多视图三维估计程序的结果获取的信息。

    Multi-view image registration with application to mosaicing and lens distortion correction
    9.
    发明授权
    Multi-view image registration with application to mosaicing and lens distortion correction 失效
    多视图图像配准应用于镶嵌和镜头失真校正

    公开(公告)号：US06173087B2

    公开(公告)日：2001-01-09

    申请号：US08966776

    申请日：1997-11-10

    申请人： Rakesh Kumar ,  Harpreet Singh Sawhney ,  James Russell Bergen

    发明人： Rakesh Kumar ,  Harpreet Singh Sawhney ,  James Russell Bergen

    IPC分类号： G06K936

    CPC分类号： G06K9/32 ,  G06K2009/2045 ,  G06T3/0081 ,  G06T3/4038 ,  H04N19/23 ,  H04N19/597

    摘要： An embodiment of the invention is a system and process for true multi-image alignment that does not rely on the measurements of a reference image being distortion free. For instance, lens distortion is a common imaging phenomenon. When lens distortion is present, none of the images can be assumed to be ideal. In an embodiment of the invention, all the images are modeled as intensity measurements represented in their respective coordinate systems, each of which is related to a reference coordinate system through an interior camera transformation and an exterior view transformation. Motion parameters determined in accordance with an embodiment of the invention dictate the position of the input frames within the reference frame. A reference coordinate system is used, but not a reference image. Motion parameters are computed to warp all input images to a virtual image mosaic in the reference coordinate system of the reference frame. Each pixel in the virtual image mosaic may be predicted by intensities at corresponding pixel positions from more than one image. The error measure, which is the sum of the variances of predicted pixel intensities at each pixel location summed over the virtual image mosaic, is minimized. The embodiment of the invention advantageously maximally uses information present in all images.

    摘要翻译： 本发明的实施例是用于真正的多图像对准的系统和过程，其不依赖于无失真的参考图像的测量。 例如，镜头失真是常见的成像现象。 当存在透镜失真时，不能假定图像是理想的。 在本发明的一个实施例中，所有图像被建模为在它们各自的坐标系中表示的强度测量值，每个坐标系统通过内部照相机变换和外部视图变换与参考坐标系相关。 根据本发明的实施例确定的运动参数指示输入框在参考框架内的位置。 使用参考坐标系，而不是参考图像。 计算运动参数以将所有输入图像扭曲成参考框架的参考坐标系中的虚拟图像马赛克。 可以通过来自多于一个图像的相应像素位置处的强度来预测虚拟图像镶嵌中的每个像素。 误差测量值是在虚拟图像镶嵌中相加的每个像素位置处的预测像素强度的方差之和。 本发明的实施例有利地最大限度地利用存在于所有图像中的信息。

    Method and apparatus for real-time pedestrian detection for urban driving
    10.
    发明授权
    Method and apparatus for real-time pedestrian detection for urban driving 有权
    城市驾驶实时行人检测方法与装置

    公开(公告)号：US08861842B2

    公开(公告)日：2014-10-14

    申请号：US13022118

    申请日：2011-02-07

    申请人： Sang-Hack Jung ,  Mayank Bansal ,  Bogdan Calin Mihai Matei ,  Jayan Eledath ,  Harpreet Singh Sawhney ,  Rakesh Kumar ,  Raia Hadsell

    发明人： Sang-Hack Jung ,  Mayank Bansal ,  Bogdan Calin Mihai Matei ,  Jayan Eledath ,  Harpreet Singh Sawhney ,  Rakesh Kumar ,  Raia Hadsell

    IPC分类号： G06K9/00

    CPC分类号： G06K9/00805 ,  G06K9/00369

    摘要： A computer implemented method for detecting the presence of one or more pedestrians in the vicinity of the vehicle is disclosed. Imagery of a scene is received from at least one image capturing device. A depth map is derived from the imagery. A plurality of pedestrian candidate regions of interest (ROIs) is detected from the depth map by matching each of the plurality of ROIs with a 3D human shape model. At least a portion of the candidate ROIs is classified by employing a cascade of classifiers tuned for a plurality of depth bands and trained on a filtered representation of data within the portion of candidate ROIs to determine whether at least one pedestrian is proximal to the vehicle.

    摘要翻译： 公开了一种用于检测车辆附近的一个或多个行人的存在的计算机实现方法。 从至少一个图像捕获装置接收场景的图像。 从图像中导出深度图。 通过将多个ROI中的每一个与3D人体形状模型进行匹配，从深度图检测多个行人候选区域（ROI）。 候选ROI的至少一部分通过采用针对多个深度带调谐的分级器级联进行分类，并且针对候选ROI部分内的数据的滤波表示进行训练，以确定至少一个行人是否靠近车辆。