专利快速检索-快速检索全球专利，免费商用专利数据库-IPRDB

1. 发明授权

US09349054B1 Foreground detector for video analytics system 有权
标题翻译：视频分析系统前景探测器
公开(公告)号：US09349054B1
公开(公告)日：2016-05-24
申请号：US14526756
申请日：2014-10-29
申请人： Behavioral Recognition Systems, Inc.
发明人： Kishor Adinath Saitwal , Lon Risinger , Wesley Kenneth Cobb
IPC分类号： G06K9/34 , G06K9/00 , G06T7/00 , G06K9/62
CPC分类号： G06K9/00785 , G06T7/11 , G06T7/136 , G06T7/194 , G06T2207/10016
摘要： Techniques are disclosed for creating a background model of a scene using both a pixel based approach and a context based approach. The combined approach provides an effective technique for segmenting scene foreground from background in frames of a video stream. Further, this approach can scale to process large numbers of camera feeds simultaneously, e.g., using parallel processing architectures, while still generating an accurate background model. Further, using both a pixel based approach and context based approach ensures that the video analytics system can effectively and efficiently respond to changes in a scene, without overly increasing computational complexity. In addition, techniques are disclosed for updating the background model, from frame-to-frame, by absorbing foreground pixels into the background model via an absorption window, and dynamically updating background/foreground thresholds.
摘要翻译：公开了用于使用基于像素的方法和基于上下文的方法来创建场景的背景模型的技术。组合方法提供了一种用于从视频流的帧中的背景分割场景前景的有效技术。此外，该方法可以缩放以同时处理大量摄像机进给，例如使用并行处理架构，同时仍然产生准确的背景模型。此外，使用基于像素的方法和基于上下文的方法确保了视频分析系统可以有效地和有效地响应场景中的变化，而不会过度增加计算复杂性。此外，公开了通过经由吸收窗口将前景像素吸收到背景模型中并且动态地更新背景/前景阈值来从帧到帧更新背景模型的技术。

2. 发明申请

US20160125621A1 INCREMENTAL UPDATE FOR BACKGROUND MODEL THRESHOLDS 有权
标题翻译：背景模型阈值的增量更新
公开(公告)号：US20160125621A1
公开(公告)日：2016-05-05
申请号：US14526879
申请日：2014-10-29
申请人： Behavioral Recognition Systems, Inc.
发明人： Kishor Adinath SAITWAL , Lon RISINGER , Wesley Kenneth COBB
IPC分类号： G06T7/40 , G06K9/38 , G06K9/00 , G06T7/20
CPC分类号： G06T7/90 , G06K9/00624 , G06K9/38 , G06K9/40 , G06T7/194 , G06T7/215 , G06T7/254 , G06T2207/10016 , G06T2207/10024 , G06T2207/20036 , G06T2207/20076
摘要： Techniques are disclosed for creating a background model of a scene using both a pixel based approach and a context based approach. The combined approach provides an effective technique for segmenting scene foreground from background in frames of a video stream. Further, this approach can scale to process large numbers of camera feeds simultaneously, e.g., using parallel processing architectures, while still generating an accurate background model. Further, using both a pixel based approach and context based approach ensures that the video analytics system can effectively and efficiently respond to changes in a scene, without overly increasing computational complexity. In addition, techniques are disclosed for updating the background model, from frame-to-frame, by absorbing foreground pixels into the background model via an absorption window, and dynamically updating background/foreground thresholds.
摘要翻译：公开了用于使用基于像素的方法和基于上下文的方法来创建场景的背景模型的技术。组合方法提供了一种用于从视频流的帧中的背景分割场景前景的有效技术。此外，该方法可以缩放以同时处理大量摄像机进给，例如使用并行处理架构，同时仍然产生准确的背景模型。此外，使用基于像素的方法和基于上下文的方法确保了视频分析系统可以有效地和有效地响应场景中的变化，而不会过度增加计算复杂性。此外，公开了通过经由吸收窗口将前景像素吸收到背景模型中并且动态地更新背景/前景阈值来从帧到帧更新背景模型的技术。

3. 发明授权

US09232140B2 Image stabilization techniques for video surveillance systems 有权
标题翻译：视频监控系统的图像稳定技术
公开(公告)号：US09232140B2
公开(公告)日：2016-01-05
申请号：US14077097
申请日：2013-11-11
申请人： Behavioral Recognition Systems, Inc.
发明人： Kishor Adinath Saitwal , Wesley Kenneth Cobb , Tao Yang
IPC分类号： H04N5/232 , G06T7/20
CPC分类号： H04N5/23267 , G06T7/246 , G06T2207/10016 , G06T2207/20021 , G06T2207/30232 , H04N5/23254 , H04N5/77
摘要： A behavioral recognition system may include both a computer vision engine and a machine learning engine configured to observe and learn patterns of behavior in video data. Certain embodiments may provide image stabilization of a video stream obtained from a camera. An image stabilization module in the behavioral recognition system obtains a reference image from the video stream. The image stabilization module identifies alignment regions within the reference image based on the regions of the image that are dense with features. Upon determining that the tracked features of a current image is out of alignment with the reference image, the image stabilization module uses the most feature dense alignment region to estimate an affine transformation matrix to apply to the entire current image to warp the image into proper alignment.
摘要翻译：行为识别系统可以包括计算机视觉引擎和被配置为观察和学习视频数据中的行为模式的机器学习引擎。某些实施例可以提供从相机获得的视频流的图像稳定性。行为识别系统中的图像稳定模块从视频流中获得参考图像。图像稳定模块基于具有特征密集的图像的区域来识别参考图像内的对准区域。在确定当前图像的跟踪特征与参考图像不对齐时，图像稳定模块使用最特征密集对准区域来估计仿射变换矩阵以应用于整个当前图像以将图像扭曲成适当的对准。

4. 发明授权

US09111148B2 Unsupervised learning of feature anomalies for a video surveillance system 有权
标题翻译：无监督学习视频监控系统的特征异常
公开(公告)号：US09111148B2
公开(公告)日：2015-08-18
申请号：US13929494
申请日：2013-06-27
申请人： BEHAVIORAL RECOGNITION SYSTEMS, Inc.
发明人： Ming-Jung Seow , Wesley Kenneth Cobb
IPC分类号： G06K9/46 , G06K9/62 , G06K9/00
CPC分类号： G06K9/00771 , G06K9/6218
摘要： Techniques are disclosed for analyzing a scene depicted in an input stream of video frames captured by a video camera. In one embodiment, e.g., a machine learning engine may include statistical engines for generating topological feature maps based on observations and a detection module for detecting feature anomalies. The statistical engines may include adaptive resonance theory (ART) networks which cluster observed position-feature characteristics. The statistical engines may further reinforce, decay, merge, and remove clusters. The detection module may calculate a rareness value relative to recurring observations and data in the ART networks. Further, the sensitivity of detection may be adjusted according to the relative importance of recently observed anomalies.
摘要翻译：公开了用于分析由摄像机捕获的视频帧的输入流中描绘的场景的技术。在一个实施例中，例如，机器学习引擎可以包括用于基于观察产生拓扑特征图的统计引擎和用于检测特征异常的检测模块。统计引擎可以包括聚类观察位置特征特征的自适应共振理论（ART）网络。统计引擎可能进一步加强，衰减，合并和删除群集。检测模块可以相对于ART网络中的重复观察和数据计算稀有度值。此外，检测的灵敏度可以根据最近观察到的异常的相对重要性进行调整。

5. 发明申请

US20150046155A1 COGNITIVE NEURO-LINGUISTIC BEHAVIOR RECOGNITION SYSTEM FOR MULTI-SENSOR DATA FUSION 有权
标题翻译：用于多传感器数据融合的认知神经元行为识别系统
公开(公告)号：US20150046155A1
公开(公告)日：2015-02-12
申请号：US14457082
申请日：2014-08-11
申请人： Behavioral Recognition Systems, Inc.
发明人： Ming-Jung SEOW , Wesley Kenneth COBB , Gang XU , Tao YANG , Aaron POFFENBERGER , Lon W. RISINGER , Kishor Adinath SAITWAL , Michael S. YANTOSCA , David M. SOLUM , Alex David HEMSATH , Dennis G. URECH , Duy Trong NGUYEN , Charles Richard MORGAN
IPC分类号： G06F17/27
CPC分类号： H04L63/1425 , G06F17/2725 , G06F17/2735 , G06F17/274 , G06F17/277 , G06F17/2775 , G06F17/2785 , G06F17/2795 , G06F17/28 , G06N99/005 , H04L63/1408
摘要： Embodiments presented herein describe techniques for generating a linguistic model of input data obtained from a data source (e.g., a video camera). According to one embodiment of the present disclosure, a sequence of symbols is generated based on an ordered stream of normalized vectors generated from the input data. A dictionary of words is generated from combinations of the ordered sequence of symbols based on a frequency at which combinations of symbols appear in the ordered sequence of symbols. A plurality of phrases is generated based an ordered sequence of words from the dictionary observed in the ordered sequence of symbols based on a frequency by which combinations of words in ordered sequence of words appear relative to one another.
摘要翻译：本文提出的实施例描述了用于生成从数据源（例如，摄像机）获得的输入数据的语言模型的技术。根据本公开的一个实施例，基于从输入数据生成的归一化矢量的有序流来生成符号序列。基于符号序列组合出现在有序序列符号中的频率，从有序符号序列的组合生成词典。基于在有序序列中的单词的组合相对于彼此出现的频率，基于在有序序列序列中观察到的字典的有序序列序列生成多个短语。

6. 发明申请

US20150003671A1 ANOMALOUS OBJECT INTERACTION DETECTION AND REPORTING 有权
标题翻译：异常对象相互作用检测和报告
公开(公告)号：US20150003671A1
公开(公告)日：2015-01-01
申请号：US13931058
申请日：2013-06-28
申请人： BEHAVIORAL RECOGNITION SYSTEMS, Inc.
发明人： Kishor Adinath SAITWAL , Dennis G. URECH , Wesley Kenneth COBB
IPC分类号： G06K9/00 , G06T7/00
CPC分类号： G06K9/00711 , G06K9/6221 , G06K9/6284
摘要： Techniques are disclosed for analyzing a scene depicted in an input stream of video frames captured by a video camera. The techniques include evaluating sequence pairs representing segments of object trajectories. Assuming the objects interact, each of the sequences of the sequence pair may be mapped to a sequence cluster of an adaptive resonance theory (ART) network. A rareness value for the pair of sequence clusters may be determined based on learned joint probabilities of sequence cluster pairs. A statistical anomaly model, which may be specific to an interaction type or general to a plurality of interaction types, is used to determine an anomaly temperature, and alerts are issued based at least on the anomaly temperature. In addition, the ART network and the statistical anomaly model are updated based on the current interaction.
摘要翻译：公开了用于分析由摄像机捕获的视频帧的输入流中描绘的场景的技术。这些技术包括评估表示物体轨迹段的序列对。假设对象相互作用，序列对中的每个序列可以被映射到自适应共振理论（ART）网络的序列簇。可以基于序列簇对的学习联合概率来确定该对序列簇的稀有度值。可以使用统计异常模型，其可以特定于交互类型或一般到多个交互类型，以确定异常温度，并且至少基于异常温度发出警报。此外，ART网络和统计异常模型基于当前的交互更新。

7. 发明申请

US20140132786A1 IMAGE STABILIZATION TECHNIQUES FOR VIDEO SURVEILLANCE SYSTEMS 有权
标题翻译：视频监控系统的图像稳定技术
公开(公告)号：US20140132786A1
公开(公告)日：2014-05-15
申请号：US14077097
申请日：2013-11-11
申请人： Behavioral Recognition Systems, Inc.
发明人： Kishor Adinath SAITWAL , Wesley Kenneth COBB , Tao YANG
IPC分类号： H04N5/232
CPC分类号： H04N5/23267 , G06T7/246 , G06T2207/10016 , G06T2207/20021 , G06T2207/30232 , H04N5/23254 , H04N5/77
摘要： A behavioral recognition system may include both a computer vision engine and a machine learning engine configured to observe and learn patterns of behavior in video data. Certain embodiments may provide image stabilization of a video stream obtained from a camera. An image stabilization module in the behavioral recognition system obtains a reference image from the video stream. The image stabilization module identifies alignment regions within the reference image based on the regions of the image that are dense with features. Upon determining that the tracked features of a current image is out of alignment with the reference image, the image stabilization module uses the most feature dense alignment region to estimate an affine transformation matrix to apply to the entire current image to warp the image into proper alignment.
摘要翻译：行为识别系统可以包括计算机视觉引擎和被配置为观察和学习视频数据中的行为模式的机器学习引擎。某些实施例可以提供从相机获得的视频流的图像稳定性。行为识别系统中的图像稳定模块从视频流中获得参考图像。图像稳定模块基于具有特征密集的图像的区域来识别参考图像内的对准区域。在确定当前图像的跟踪特征与参考图像不对齐时，图像稳定模块使用最特征密集对准区域来估计仿射变换矩阵以应用于整个当前图像以将图像扭曲成适当的对准。

8. 发明申请

US20140003710A1 UNSUPERVISED LEARNING OF FEATURE ANOMALIES FOR A VIDEO SURVEILLANCE SYSTEM 有权
标题翻译：视频监控系统特征异常的不间断学习
公开(公告)号：US20140003710A1
公开(公告)日：2014-01-02
申请号：US13929494
申请日：2013-06-27
申请人： BEHAVIORAL RECOGNITION SYSTEMS, Inc.
发明人： Ming-Jung SEOW , Wesley Kenneth COBB
IPC分类号： G06K9/00
CPC分类号： G06K9/00771 , G06K9/6218
摘要： Techniques are disclosed for analyzing a scene depicted in an input stream of video frames captured by a video camera. In one embodiment, e.g., a machine learning engine may include statistical engines for generating topological feature maps based on observations and a detection module for detecting feature anomalies. The statistical engines may include adaptive resonance theory (ART) networks which cluster observed position-feature characteristics. The statistical engines may further reinforce, decay, merge, and remove clusters. The detection module may calculate a rareness value relative to recurring observations and data in the ART networks. Further, the sensitivity of detection may be adjusted according to the relative importance of recently observed anomalies.
摘要翻译：公开了用于分析由摄像机捕获的视频帧的输入流中描绘的场景的技术。在一个实施例中，例如，机器学习引擎可以包括用于基于观察产生拓扑特征图的统计引擎和用于检测特征异常的检测模块。统计引擎可以包括聚类观察位置特征特征的自适应共振理论（ART）网络。统计引擎可能进一步加强，衰减，合并和删除群集。检测模块可以相对于ART网络中的重复观察和数据计算稀有度值。此外，检测的灵敏度可以根据最近观察到的异常的相对重要性进行调整。

9. 发明授权

US08548198B2 Identifying anomalous object types during classification 有权
标题翻译：在分类期间识别异常对象类型
公开(公告)号：US08548198B2
公开(公告)日：2013-10-01
申请号：US13622281
申请日：2012-09-18
申请人： Behavioral Recognition Systems, Inc.
发明人： Wesley Kenneth Cobb , David Friedlander , Rajkiran Kumar Gottumukkal , Ming-Jung Seow , Gang Xu
IPC分类号： G06K9/00 , G01V3/00
CPC分类号： G06K9/6251 , G06K9/00771
摘要： Techniques are disclosed for identifying anomaly object types during classification of foreground objects extracted from image data. A self-organizing map and adaptive resonance theory (SOM-ART) network is used to discover object type clusters and classify objects depicted in the image data based on pixel-level micro-features that are extracted from the image data. Importantly, the discovery of the object type clusters is unsupervised, i.e., performed independent of any training data that defines particular objects, allowing a behavior-recognition system to forgo a training phase and for object classification to proceed without being constrained by specific object definitions. The SOM-ART network is adaptive and able to learn while discovering the object type clusters and classifying objects and identifying anomaly object types.
摘要翻译：公开了用于在从图像数据提取的前景对象的分类期间识别异常对象类型的技术。基于从图像数据提取的像素级微特征，使用自组织图和自适应共振理论（SOM-ART）网络来发现对象类型簇并对图像数据中描绘的对象进行分类。重要的是，对象类型簇的发现是无监督的，即独立于定义特定对象的任何训练数据执行，允许行为识别系统放弃训练阶段，并且对象分类进行而不受特定对象定义的约束。 SOM-ART网络是自适应的，能够在发现对象类型集群并分类对象并识别异常对象类型时学习。

10. 发明申请

US20130136353A1 BACKGROUND MODEL FOR COMPLEX AND DYNAMIC SCENES 审中-公开
标题翻译：复杂和动态场景的背景模型
公开(公告)号：US20130136353A1
公开(公告)日：2013-05-30
申请号：US13746760
申请日：2013-01-22
申请人： Behavioral Recognition Systems, Inc.
发明人： Wesley Kenneth COBB , Ming-Jung SEOW , Tao YANG
IPC分类号： G06K9/62
CPC分类号： G06T7/20 , G06K9/00335 , G06K9/00718 , G06K9/00771 , G06K9/6218 , G06K9/6222 , G06K9/66 , G06K2009/00738 , G06T7/254 , G06T2207/10016 , G06T2207/20084 , G06T2207/30196 , G06T2207/30232
摘要： Techniques are disclosed for learning and modeling a background for a complex and/or dynamic scene over a period of observations without supervision. A background/foreground component of a computer vision engine may be configured to model a scene using an array of ART networks. The ART networks learn the regularity and periodicity of the scene by observing the scene over a period of time. Thus, the ART networks allow the computer vision engine to model complex and dynamic scene backgrounds in video.
摘要翻译：披露了一种技术，用于在无监督的一段时间内为复杂和/或动态场景学习和建模背景。计算机视觉引擎的背景/前景分量可以被配置为使用ART网络阵列对场景建模。 ART网络通过观察一段时间内的场景，了解场景的规律性和周期性。因此，ART网络允许计算机视觉引擎在视频中对复杂和动态的场景背景进行建模。

你已经成功收藏专利！

检索式保存成功!

IPRDB

热门服务

关于我们

友情链接

联系方式