Long exposure video-surveillance : isolation of new object on the scenario and rejection of detection due to movement of background objects

Repositorio e-Archivo

Mostrar el registro sencillo del ítem

dc.contributor.advisor Caglioti, Vincenzo
dc.contributor.advisor Carpintero del Barrio, Guillermo (tutor)
dc.contributor.author Gómez de Dueñas, Sandra
dc.contributor.other Politecnico di Milano. Dipartimento di Elettronica e Informazione
dc.date.accessioned 2011-12-26T11:08:12Z
dc.date.available 2011-12-26T11:08:12Z
dc.date.issued 2010
dc.date.submitted 2011-10-03
dc.identifier.uri http://hdl.handle.net/10016/12844
dc.description.abstract Video-surveillance is getting more and more importance nowadays. The main target is improving the level of understanding of the scene analyzed. However, when the level of scene detail is pushed into the background and the main objective of the system is activity detection, the low-level technique presented here could be, theoretically, more effective. When using long-exposure images of the same scene, objects in motion appears blurred and semi-transparent, and these features increase with the time of exposure, until being totally invisible when using large times. As the main context where it is applied is Wireless Sensor Networks or similar, which require the minimal computing and communication features, using long-exposure images produces an important decrease in the number of frames to send to the surveillance-station, and linked to the technique’s low-computational effort required, both facts enlarge the lifetime of the sensor batteries. The basic method implemented in order to detect activity is differencing two consecutive images. The difference image represents the state of motion which has taken place during the lapse-time of both photographies, i.e., (theoretically), a black image if no activity is detected and, on the contrary, if an object has entered in the scene or a motion is produced on the scene the difference image pixels containing information about the motion will be represented in a colour, different from black. This is much easier of understanding if the difference image is later transformed into a “black and white image”. Then, if no motion has been produced, a (theoretically) the difference image should be totally black, but if instead, a motion is produced the pixels affected by motion will be represented in white. Consequently, the device operates in two mainly modes 1)Idle Mode: It is expected that the great part of the time the device works in this mode, in which, it is waiting for any new activity on the scene. 2)Alert Mode: Once a presence is detected, the network is switched into this mode,and it should send an alert, and a new detection state, with more complex and specific detection and analyzing resources should be executed.
dc.format.mimetype application/pdf
dc.format.mimetype application/pdf
dc.language.iso eng
dc.rights Atribución-NoComercial-SinDerivadas 3.0 España
dc.rights.uri http://creativecommons.org/licenses/by-nc-nd/3.0/es/
dc.subject.other Proceso de imágenes
dc.subject.other Vídeo
dc.title Long exposure video-surveillance : isolation of new object on the scenario and rejection of detection due to movement of background objects
dc.type masterThesis
dc.subject.eciencia Electrónica
dc.rights.accessRights openAccess
dc.description.degree Ingeniería Industrial
dc.contributor.departamento Universidad Carlos III de Madrid. Departamento de Tecnología Electrónica
 Find Full text

Ficheros en el ítem

*Click en la imagen del fichero para previsualizar.(Los elementos embargados carecen de esta funcionalidad)

El ítem tiene asociada la siguiente licencia:

Este ítem aparece en la(s) siguiente(s) colección(es)

Mostrar el registro sencillo del ítem