多旋翼鲁棒三维视觉跟踪

　　There are 7 chapters in this book. Chapter 1 introduces the three-dimensional（3D）visual tracking problem for multicopters. Chapter 2 gives some preliminaries used in the book. Chapters 3~4 focus on the one-dimensional（1D） calibration problem for multiple perspective cameras and multiple fish-eye cameras. Chapters 5~6 deal with the robust pose estimation problem of multicopters in single-camera case and multi-camera case. Chapter 7 introduces a 3D visual tracking system for multicopters built with cheap and off-the-shelf cameras.
　　This book is written for researchers and engineers working in robotic visual servo-ing， machine vision， systems and control， etc.

　　Nowadays multicopters have been widely used in many applications，such as aerial photography，goods transportation and so on. Vision-based pose esti-mation （or three-dimensional （3D）tracking）of these vehicles is a crucial and interesting problem for their autonomous operation. The 3D visual track-ing system for multicopters is widely utilized because it could provide accu-rate six-degree-of-freedom pose of single or multiple multicopters，and serve as a testbed to provide a quick navigation solution for testing and evalu-ating flight control and guidance algorithms.
　　This book focuses on the two key technologies in a 3D visual tracking sys-tem for multicopters：（1）Multi-camera calibration；（2）Pose estimation for multicopters. It presents new methods for calibrating multiple perspective cameras and multiple fish-eye cameras based on a freely-moving one-di-mensional （1D） calibration object. This book also gives new pose estimation methods for multicopters based on the process model for multicopters and sin-gle or multiple off-board cameras.
　　Tlus book contains both theoretical results and practical experiments. It is written for researchers and engineers working in robotic visual servoing，ma-chine vision，systems and control，etc. It can also be used as a supplemen-tary reading material for senior undergraduate students and graduate students in the areas of utomation，robotics，artificial intelligence，etc.
　　This book could not be finished without the help and encouragement of many people. We give our thanks to the former colleagues in Beihang Univer-sity for substantial support. We also want to thank the colleagues in Universityof Science and Technology Beijing and many other researchers for providing valuable comments and suggestions.This book is supported by the National Natural Science Foundation of China under Grant 61803025.