Design And Implementation Of Outdoor Autonomous Searching Aircraft

As the outdoor sports are becoming more and more popular in recent years,the events of getting lost in wild happen frequently.To solve the problem of searching and rescuing outdoors,the methods of ground-based searching and observing by people in helicopter are usually used in our country.However,these methods are limited by the complex environment,especially at night.An autonomous searching aircraft(ASA)has been designed and made in this paper to solve the problem of searching and rescuing in wild.The ASA adopts strap-down inertial navigation system(SINS)to calculate the attitude,fuses the data of barometer and the GPS to estimate height,besides the Cascade PID control algorithm is used to control pose and position.The ASA can plan trajectories autonomously according to the areas set by ground station,and the searching and rescuing task could be implemented independently and efficiently.Infrared images in searching areas are collected by the infrared imager named FLIR TAU2 which is mounted the aircraft.Then the images will be processed by the ODROID-XU4 to realize real-time detection.By sending the position of the trapped person to the ground station,the search team can complete the task efficiently.In the aspect of infrared detection,the infrared target is described by histogram of oriented gradient(HOG)feature;the classification and detection of HOG feature are completed by the linear support vector machine(SVM)classifier.At the same time,the SVM has been trained by self-built training set to meet the need of infrared human target detection.Experiment results show that the algorithm designed in this project has simple structure,strong effectiveness and the ASA designed in this project can implement the autonomous searching and rescuing task in the wild.