Under CopyrightGrenz, CarstenCarstenGrenzJänen, UweUweJänenHähner, JörgJörgHähnerKuntzsch, ColinColinKuntzschMenze, MoritzMoritzMenzeD'Angelo, DavidDavidD'AngeloBogen, ManfredManfredBogenMonari, EduardoEduardoMonari2022-03-1216.1.20132012https://publica.fraunhofer.de/handle/publica/37753710.24406/publica-fhg-377537The CamInSens system is a next-generation selforganizing video surveillance system that combines research being done in the fields of person-tracking, trajectory analysis, visual analytics, and self-organizing system management algorithms. Its purpose is the online threat detection by analysing anomalies in persons trajectories. Therefore, robust multicamera multi-person tracking is combined with a flexible analysis module, which uses online learning classification algorithms as well as user-generated filters to process the persons trajectories in the surveillance space.endistributed resource managementmulti-camera multi-person trackingself-learning classification of spatio-temporal trajectory datavisualization and interaction with self-organizing smart camera systemsclassification algorithmdistributed smart camerasmulti-person trackingself-organizing systemthreat detectiontrajectory analysisvideo surveillance systemsvisual analytics005006629004670conference paper