Desarrollo de un sistema de visión artificial para el seguimiento de objetos en el espacio y determinación de sus trayectorias.

Abstract

Artificial vision is a discipline that is being studied in recent decades. Artificial vision systems have been used with great success in industrial processes such as automation, inspection or quality control, surveillance or security systems, medical processes, etc. Its objectives and advantages range from the improvement in the speed of the processes to the automation of tasks that cannot be done otherwise. For this reason, the development of a system for object tracking, location and trajectory generation in a three-dimensional space in real time is presented. This system is developed in the python programming language, using computer vision techniques and artificial intelligence, studying different tools, techniques and algorithms from which the best algorithms are selected and adapted. From the base of methods, develop the algorithms that specified the solution to the established requirements. These algorithms are embedded in a functional prototype that facilitates their operation through a graphical interface that allows the configuration of the tracking camera system. All algorithms are fully documented with didactic fines. The system was evaluated in different controlled scenarios where the tracking error, location and trajectory generation is determined.