Implementation of an Adaptive Artificial Vision System Applied to Video Surveillance Tasks on Raspberry Pi

Video surveillance tools and systems are a fundamental part of our current society, even more so when there is constant growth both in the technological aspect and in the multiple needs linked to the development of intelligent cities and the various areas of the commercial and productive sectors of a country. This work proposes the implementation of an adaptive artificial vision system applied to video surveillance tasks, making use of a Raspberry PI development card and the OpenCV image processing libraries. The aim is to integrate an adaptive background subtraction algorithm based on the calculation of the median technique as a fundamental basis for the detection and tracking of objects, both in outdoor and indoor environments. The aim of the work is to present a scalable software tool that can be adapted to any type of domotic or IoT application, taking advantage of the multiple expansion possibilities that low-cost development cards such as Raspberry have.