Desarrollo e implementación de un sistema de visión artificial y seguimiento de objetivos humanos por un cuadricóptero de exteriores utilizando Matlab

Abstract

In the present project an artificial vision system was developed, with the main purpose of demonstrating that its application is possible in the detection of human faces, in real time, through a video input. By means of image processing the control signals were obtained, in order to controlling the drone in real time. To achieve this goal, this project was carried out in stages. Considering that the detection methodology was a cascade detector using the algorithm of Viola Jones. Initially, the equipment and software necessary, for the acquisition of the video captured by the camera, were selected. Later, an image processing system was developed, in order to improve the quality of the image and prepare it for detection. Then, through the segmentation of colors of each frame, thresholding, binarization and edge detection, the program was able to automatically define a region for each image in which the human objective is located. In this way, having the region where the face is most likely to be found, a cascade classifier is trained, which through a tracking algorithm, will allow to obtain the position and area where the face is located from the center of the image. Finally, using a controller connected to the drone, the flight trajectory will be controlled in order to keep it attached to the target.