Research Of Object Detection System For MINI UAV

Precise maneuverability of the Mini unmanned helicopter makes them useful for many critical tasks ranging from rescue and security to inspection and aerial mapping. Therefore, the Mini unmanned helicopter systems have recently attracted considerable attention. For instance, more and more universities and institutes enter the international aerial robotics competition to show their new advanced UAV systems annually.Image processing system is a very important part of Mini unmanned helicopter, which mainly dealing with how to detect the object and supply the navigation computer with location information of the object. However, the varied background and the moving of the camera make the image processing become difficult. To solve these problems, this thesis introduces an object detection algorithm based on contour retrieving, template match and HU moments invariant. A cascade classifier based on HAAR like features and ADABOOST algorithm is compared with the former. Using these two object detection algorithms, a real-time image processing system for UAV is constructed. The frame rate of the two algorithms both can reach 6-7 Hz. In the end, the experimental results show the system effectiveness.