Sistema prototipo de detección de movimiento y tracking de personas con control sobre el pan-tilt de la cámara como apoyo a sistemas de videovigilancia

Abstract

This document exposes and highlights the importance of image processing techniques, development and implementation of controllers, installation and start-up of projects aimed at video surveillance or security systems, whose value lies in guaranteeing the safety of the person to be detected and on other occasions, the integrity of the goods and facilities, additionally they are used for research purposes, health care or the control of labor provision by workers. And it is due to the constant change in the hardware, techniques and data management in the detection / tracking of objectives, in how much they are on par with the photographic and video cameras available on the market, which make this project an ideal reference. current technology and the contributions that can be made to the systems implemented in the nation. A significant set of video surveillance applications pose static camera systems, which translates into detection errors in the background change (environment through which the objectives are mobilized), movement outside the range captured by the camera, or by abrupt changes in the system that cause errors in the pursuit of one or more objectives. Reason why it is convenient in this type of systems to modify the location of each camera to areas of interest, or where there is a greater number of people or objectives, this last approach is established in methodologies that are directed to vehicles (manned or not ), robots, part of industrial machinery among other security systems or subsystems. Then it is possible to give a change to the implementations by static camera, using a widening of the field of view (FOV), emphasized in the project with a movement in the pan and tilt axis of the camera, of course contemplating disturbances by lighting, changes of scene (result of the same movement of the camera), communication, among others, in addition to the focus that allows to determine a reference value or of "something" that compensates it, then the developed applies to the maintenance of the optical axis of the camera in the main frame (frame where the transmission of what is captured by the image sensor is seen). The mapping of said frame serves to find relevant information on the behavior of the system by feedback control, and of course the error signal associated with the set point that varies according to the video source. The result is the follow-up of a person (s) found in the FOV of the main frame, with the detection and adjustment algorithm carried out by the system with the developed control strategy, secondly with the indications that make it pertinent to draw a route of the objective by means of simulation tools; After how, it may differ from a relatively acceptable operation, given by the error signal established in the main algorithm and finally by the validation of the operability by means of a prototype of the system.