Agt International Gmbh

Ulf Blanke

Markus Schlattmann

Henning Hamer

Holger Ziekow

Tim Van Kasteren

Martin Berchtold

Alessandro Leonardi

Jelena Mirkovic

Maximilian Walther

A computer system, method, and program product for registering sensors, includes an interface to receive a query (Q) related to a specific event for a target system state associated with a specific activity pattern. A decomposer decomposes the query (Q) into semantic expressions (QSe,QSe) describing the target system state, and identifies a database entry associated with a semantic expression from the decomposition. A retriever retrieves behavioral data corresponding to the identified database entry, including dependencies between elements that describe the specific activity pattern. The specific activity pattern is associated with multiple activity data structures. A registration component registers a sensor (S,S,S), assigned to the activity data structures.

System and method for registering sensors used in monitoring-systems

In queues, persons&#x2014;or objects () in general&#x2014;move inside an area () to a target (), such as to a counter. The queue has movement characteristics in terms of speed, waiting times and queue form. A computer-implemented approach obtains the characteristics by receiving a sequence () of image frames () that represent the surveillance area (); calculating flow vectors that indicate an optical displacement for the sequence () of image frames (); extending one of the flow vectors as lead vector in extension directions; determining intermediate vectors by using flow vectors along the extension directions; and selecting one the intermediate vectors as the new lead vector. The steps are repeated to concatenate the lead vectors to the movement characteristics ().

Image processing to derive movement characteristics for a plurality of queue objects

A computer implemented method, computer program product, and computer system for determining camera calibration data. The computer system receives geo-positional data of a moving object, wherein the geo-positional data is associated with an indicator (). The computer system receives further a sequence of frames () from the at least one camera (), wherein at least one frame has a picture of the moving object () with a structure and with an encoded version of the indicator which are optically recognizable. The indicator associated with the at least one frame is extracted by decoding () the optically encoded version of the indicator of the at least one frame. The geo-positional data of the moving object which is in the picture of the at least one frame is obtained by matching () the indicator associated with the geo-positional data of the moving object and the decoded indicator associated with the at least one frame. At least one reference point of the at least one frame is identified by analyzing the optically recognizable structure in the picture of the at least one frame. The camera calibration data of the at least one camera is determined by applying a geometric transformation () on the at least one reference point and its associated geopositional data of the moving object which is in the at least one frame.

Calibration of camera-based surveillance systems

Growing community of inventors

Zurich, Switzerland

Ulf Blanke