Rastreamento de pessoas em sequência de imagens infravermelhas

Abstract

This paper approaches the problem of tracking certain individuals in infrared images from previously selected environments. This project was developed using the method threshold to detect objects aiming to separate wanted objects from inputted images. Using the earlier discovered objects, condensation algorithm was applied in order to determine the path of the objects over time. The inputted videos were recorded by a camera which was capable of recording infrared images. By applying algorithms of tracking on the recorded images we will be able to measure the temperature of each individual in the environment in noninvasive ways. Furthermore, by using the tracking algorithms over time it is possible to obtain a log of information about the wanted objects. Additionally, the temperature will be measured a bunch of times in order to improve the trust of the algorithms.