SLAM for drones : simultaneous localization and mapping for autonomous flying robots

The main objective of this thesis is to be a reference in SLAM for future work in robotics. It goes from almost a zero-point for a non-expert in the field until a revision of the SoA methods. It has been carefully divided into four parts: - The first one is a compilation of the basis in computer vision. If you are new into the field, it is recommended to read it carefully to really understand the most important concepts that will be applied in further sections. - The second part will be a full revision from zero of SLAM techniques, focusing on the award winning KinectFusion and other SoA methods. - The third part goes from a general flying robots overview in history until the mechanical model of a quadrotor. It has been intended to be completely apart from section two, for the case it has been determined to only focus on the vision part of this thesis. - The fourth part is a pro-cons overview of the SLAM methods described, applied into flying robots. We will finish with the conclusions and future work of this MSc research. ____________________________________________________________________________________________________________________El objetivo del proyecto es realizar documento que ordene, clasifique y explique desde un nivel básico hasta las técnicas más punteras, todo lo que el acrónimo SLAM engloba. Además nos focalizaremos en concreto en resolver el problema para robots voladores no tripulados. El documento original se divide en cuatro bloques principales precedidos por agradecimientos, una definición de los objetivos de la tesis, e introducción. Éstos cuatro bloques son: 1: Primera parte: Conceptos básicos de visión por computador 2: Segunda parte: S.L.A.M. 3: Tercera parte: cuadrotores 4: Cuarta parte: SLAM para robots voladores Por último incluye un apartado de trabajo futuro y conclusiones.Ingeniería Industria