公开(公告)号：US20180181867A1

公开(公告)日：2018-06-28

申请号：US15851173

申请日：2017-12-21

Applicant: AXIS AB

Inventor： Robin Seibold ,  Jiandan Chen ,  Hanna Björgvinsdóttir ,  Martin Ljungqvist

IPC: G06N3/08 ,  G06K9/62

Abstract: Methods and apparatus, including computer program products, implementing and using techniques for configuring an artificial neural network to a particular surveillance situation. A number of object classes characteristic for the surveillance situation are selected. The object classes form a subset of the total number of object classes for which the artificial neural network is trained. A database is accessed that includes activation frequency values for the neurons within the artificial neural network. The activation frequency values are a function of the object class. Those neurons having activation frequency values lower than a threshold value for the subset of selected object classes are removed from the artificial neural network.

公开(公告)号：US09830528B2

公开(公告)日：2017-11-28

申请号：US14963792

申请日：2015-12-09

Applicant: Axis AB

Inventor： Jiandan Chen ,  Anders Lloyd ,  Niclas Danielsson

IPC: G06K9/46 ,  G06K9/62

CPC classification number: G06K9/4647 ,  G06K9/46 ,  G06K9/6202

Abstract: A method may include determining a value indicative of an average intensity of blocks in an image. The blocks include a primary and outer blocks. Each of the outer blocks may have three, five, or more than five pixels. The image may describe an external pixel lying between the primary and at least one of the outer blocks. The external pixel may not contribute to the value indicative of the average intensity of any of the blocks. The image may also describe a common internal pixel lying within two of the blocks. The common pixel may contribute to the value indicative of the average intensity of the two of the blocks. The method may include comparing the value indicative of the average intensity of the primary block to the values of the outer blocks, and quantifying a feature represented by the image by generating a characteristic number.

公开(公告)号：US09536154B2

公开(公告)日：2017-01-03

申请号：US14273181

申请日：2014-05-08

Applicant: Axis AB

Inventor： Markus Skans ,  Anders Johannesson ,  Jiandan Chen ,  Joakim Baltsen

IPC: G06K9/00 ,  G08B13/196 ,  G06K9/20 ,  G06K9/62 ,  G06T7/20

CPC classification number: G06K9/00771 ,  G06K9/00369 ,  G06K9/2036 ,  G06K9/2054 ,  G06K9/6201 ,  G06T7/292 ,  G06T2207/10016 ,  G06T2207/20112 ,  G06T2207/30196 ,  G06T2207/30232 ,  G08B13/19652 ,  G08B13/19686

Abstract: A method of monitoring a scene by a camera (7) comprises marking a part (14) of the scene with light having a predefined spectral content and a spatial verification pattern. An analysis image is captured of the scene by a sensor sensitive to the predefined spectral content. The analysis image is segmented based on the predefined spectral content, to find a candidate image region. A spatial pattern is detected in the candidate image region, and a characteristic of the detected spatial pattern is compared to a corresponding characteristic of the spatial verification pattern. If the characteristics match, the candidate image region is identified as a verified image region corresponding to the marked part (14) of the scene. Image data representing the scene is obtained, and image data corresponding to the verified image region is processed in a first manner, and remaining image data is processed in a second manner.

Abstract translation: 通过照相机（7）监视场景的方法包括用具有预定义频谱内容和空间验证模式的光标记场景的一部分（14）。 通过对预定义的光谱内容敏感的传感器捕获场景的分析图像。 分析图像基于预定义的光谱内容被分割，以找到候选图像区域。 在候选图像区域中检测空间图案，并将检测到的空间图案的特性与空间验证图案的相应特征进行比较。 如果特征匹配，则候选图像区域被识别为对应于场景的标记部分（14）的验证图像区域。 获得表示场景的图像数据，并且以第一方式处理与经验证的图像区域相对应的图像数据，并且以第二方式处理剩余的图像数据。