Realization of Scenarios for Video Surveillance

Abstract

The design of a methodology for the effective scene understanding systems is one of the main goals of the researchers in the analysis of video surveillance. The objects in the scene have to be identified. Hence, it is necessary to detect the parts belonging to the background. In the article we introduce the base algorithms, which enable us to realization of scenarios. We briefly describe base algorithms (object detection, object localization, recognition of humans, movement detection and configuration of scene) used in three selected scenarios: violation of protected zones, abandoned objects and vandalism (graffiti). These scenarios were tested on several films, obtained from Internet and made by participants of project SIMPOZ. The results of our experiments are presented. The basic algorithms for detecting and locating objects are very quickly, but movement detection ("optical flow") and recognition of humans algorithms work longer.

Risk assessment for a video surveillance system based on Fuzzy Cognitive Maps

“…Video detection components within SIMPOZ system provide several monitoring functions [14]: violation of protected zones, detection of movement in a forbidden direction, object abandonment, theft, loitering, crowd gathering, vandalism (graffiti and devastation) and fight. Trajectory collision detection is a special feature dedicated to protection of such facilities as airports.…”

Object detection and tracking for low-cost video surveillance system

“…The algorithm follows an overall scheme of the video surveillance system SIMPOZ described in [4] and introduces details of implementation of particular modules and some modifications of the general scheme. Additionally, the executive module has been introduced which is responsible for performing actions like email notification, alarm generation, system update, watchdog... .…”

“…Researchers present various parts of the system, ranging from individual components (i.e. the background generation [9]) to overall systems that can interpret and recognize various scenarios in automated surveillance (SIMPOZ [4]). A comprehensive review of background modeling for foreground detection has been given in article [3].…”

A State-Based Language for Enhanced Video Surveillance Modeling (SEL)

“…One of the many advantages of using a grammatical method is that it can leverage current data and reduce duplicates, which can lead to the eventual elimination of the requirement for redundant data collecting or extra training [16,17]. Although using formal language to describe and model activities provides remarkable expressiveness, it poses significant challenges when interpreting and translating data from real-world scenarios.…”