公开(公告)号：US08649594B1

公开(公告)日：2014-02-11

申请号：US12802265

申请日：2010-06-03

申请人： Wei Hua ,  Juwei Lu ,  Jinman Kang ,  Jon Cook ,  Haisong Gu

发明人： Wei Hua ,  Juwei Lu ,  Jinman Kang ,  Jon Cook ,  Haisong Gu

IPC分类号： G06K9/62

CPC分类号： G06K9/00771 ,  G06K9/6263

摘要： A method for assessing events detected by a surveillance system includes assessing the likelihood that the events correspond to events being monitored from feedback in response to a condition set by a user. Classifiers are created for the events from the feedback. The classifiers are applied to allow the surveillance system improve its accuracy when processing new video data.

摘要翻译： 用于评估由监视系统检测到的事件的方法包括响应于由用户设置的条件，从反馈评估事件对应于正在监视的事件的可能性。 针对来自反馈的事件创建分类器。 应用分类器来允许监视系统在处理新的视频数据时提高其准确性。

公开(公告)号：US20140029788A1

公开(公告)日：2014-01-30

申请号：US13558538

申请日：2012-07-26

申请人： Jinman Kang

发明人： Jinman Kang

IPC分类号： G06K9/36

CPC分类号： G06K9/2018 ,  G06K9/0051

摘要： Detecting an object includes receiving depth data and infrared (IR) data from a depth sensor. A first background subtraction is performed on the IR data to create a first mask, and a second background subtraction is performed on the IR data to create a second mask. The first and second masks and the depth data are merged to create a third mask.

摘要翻译： 检测对象包括从深度传感器接收深度数据和红外（IR）数据。 对IR数据执行第一背景减除以创建第一掩码，并且对IR数据执行第二背景扣除以创建第二掩码。 第一和第二掩码和深度数据被合并以创建第三掩码。

公开(公告)号：US20130300659A1

公开(公告)日：2013-11-14

申请号：US13470885

申请日：2012-05-14

申请人： Jinman Kang ,  Charles Jia

发明人： Jinman Kang ,  Charles Jia

IPC分类号： G06F3/033

CPC分类号： G06F3/017

摘要： Recognizing a command may include monitoring a tangible reference with a depth sensor, maintaining a virtual reference approximately on calibrated three dimensional coordinates of the tangible reference, maintaining a touch space adjacent the virtual reference, and recognizing a command when a predetermined object enters the touch space.

摘要翻译： 识别命令可以包括使用深度传感器来监测有形参考，将虚拟参考大致保持在有形参考的校准三维坐标上，保持与虚拟参考相邻的触摸空间，以及当预定对象进入触摸空间时识别命令 。

公开(公告)号：US20080273751A1

公开(公告)日：2008-11-06

申请号：US11873390

申请日：2007-10-16

申请人： Chang Yuan ,  Gerard Medioni ,  Jinman Kang ,  Isaac Cohen

发明人： Chang Yuan ,  Gerard Medioni ,  Jinman Kang ,  Isaac Cohen

IPC分类号： G06K9/00

CPC分类号： G06K9/32 ,  G06T7/215 ,  G06T7/579

摘要： Among other things, methods, systems and computer program products are described for detecting and tracking a moving object in a scene. One or more residual pixels are identified from video data. At least two geometric constraints are applied to the identified one or more residual pixels. A disparity of the one or more residual pixels to the applied at least two geometric constraints is calculated. Based on the detected disparity, the one or more residual pixels are classified as belonging to parallax or independent motion and the parallax classified residual pixels are filtered. Further, a moving object is tracked in the video data. Tracking the object includes representing the detected disparity in probabilistic likelihood models. Tracking the object also includes accumulating the probabilistic likelihood models within a number of frames during the parallax filtering. Further, tracking the object includes based on the accumulated probabilistic likelihood models, extracting an optimal path of the moving object.

摘要翻译： 其中，描述了用于检测和跟踪场景中的移动物体的方法，系统和计算机程序产品。 从视频数据识别一个或多个残留像素。 至少两个几何约束被应用于所识别的一个或多个残余像素。 计算一个或多个残余像素与应用的至少两个几何约束的差异。 基于检测到的视差，将一个或多个残余像素分类为属于视差或独立运动，并且对视差分类的残留像素进行滤波。 此外，在视频数据中跟踪移动对象。 跟踪对象包括表示在概率似然模型中检测到的差异。 跟踪对象还包括在视差滤波期间在多个帧内累积概率似然模型。 此外，跟踪对象包括基于累积的概率似然模型，提取移动对象的最佳路径。

公开(公告)号：US08831285B2

公开(公告)日：2014-09-09

申请号：US13558538

申请日：2012-07-26

申请人： Jinman Kang

发明人： Jinman Kang

IPC分类号： G06K9/00 ,  H04N9/67

CPC分类号： G06K9/2018 ,  G06K9/0051

摘要： Detecting an object includes receiving depth data and infrared (IR) data from a depth sensor. A first background subtraction is performed on the IR data to create a first mask, and a second background subtraction is performed on the IR data to create a second mask. The first and second masks and the depth data are merged to create a third mask.

摘要翻译： 检测对象包括从深度传感器接收深度数据和红外（IR）数据。 对IR数据执行第一背景减除以创建第一掩码，并且对IR数据执行第二背景扣除以创建第二掩码。 第一和第二掩码和深度数据被合并以创建第三掩码。

公开(公告)号：US08073196B2

公开(公告)日：2011-12-06

申请号：US11873390

申请日：2007-10-16

申请人： Chang Yuan ,  Gerard Medioni ,  Jinman Kang ,  Isaac Cohen

发明人： Chang Yuan ,  Gerard Medioni ,  Jinman Kang ,  Isaac Cohen

IPC分类号： G06K9/00

CPC分类号： G06K9/32 ,  G06T7/215 ,  G06T7/579

摘要： Among other things, methods, systems and computer program products are described for detecting and tracking a moving object in a scene. One or more residual pixels are identified from video data. At least two geometric constraints are applied to the identified one or more residual pixels. A disparity of the one or more residual pixels to the applied at least two geometric constraints is calculated. Based on the detected disparity, the one or more residual pixels are classified as belonging to parallax or independent motion and the parallax classified residual pixels are filtered. Further, a moving object is tracked in the video data. Tracking the object includes representing the detected disparity in probabilistic likelihood models. Tracking the object also includes accumulating the probabilistic likelihood models within a number of frames during the parallax filtering. Further, tracking the object includes based on the accumulated probabilistic likelihood models, extracting an optimal path of the moving object.

摘要翻译： 其中，描述了用于检测和跟踪场景中的移动物体的方法，系统和计算机程序产品。 从视频数据识别一个或多个残留像素。 至少两个几何约束被应用于所识别的一个或多个残余像素。 计算一个或多个残余像素与应用的至少两个几何约束的差异。 基于检测到的视差，将一个或多个残余像素分类为属于视差或独立运动，并且对视差分类的残留像素进行滤波。 此外，在视频数据中跟踪移动对象。 跟踪对象包括表示在概率似然模型中检测到的差异。 跟踪对象还包括在视差滤波期间在多个帧内累积概率似然模型。 此外，跟踪对象包括基于累积的概率似然模型，提取移动对象的最佳路径。

公开(公告)号：US09652043B2

公开(公告)日：2017-05-16

申请号：US13470885

申请日：2012-05-14

申请人： Jinman Kang ,  Charles Jia

发明人： Jinman Kang ,  Charles Jia

IPC分类号： G06F3/033 ,  G06F3/01

CPC分类号： G06F3/017

摘要： Recognizing a command may include monitoring a tangible reference with a depth sensor, maintaining a virtual reference approximately on calibrated three dimensional coordinates of the tangible reference, maintaining a touch space adjacent the virtual reference, and recognizing a command when a predetermined object enters the touch space.