基于视觉引导的无人直升机着舰技术研究

以无人直升机为被控对象,根据实际着舰流程研究了视觉引导系统的软硬件架构。在硬件上通过在摄像头加装云台,避免了无人机因姿态改变造成视野中目标图像的丢失,提高了位姿解算的精度。提出一种以红外灯设计的合作目标图案为着舰目标的视觉引导着舰系统,能够在不同光线强度、复杂周边环境下成功识别合作目标,具有良好的抗噪能力。研究设计了一套图像处理与位姿解算的引导系统,通过实验验证姿态参数的误差在2°以内,位置参数的误差小于2cm,能够满足着舰的精度要求;每帧图像的处理速度在30ms左右,具有良好的实时性。

Research on Landing Technology of Unmanned Helicopter Based on Vision Guidance

Based on the actual landing process of unmanned helicopter, the hardware and software architecture of the visual guidance system is studied. By installing the camera under the pan-tilt-zoom, the loss of the target image in the field of vision caused by the change of attitude of the UAV can be avoided, and the accuracy of position and attitude can be improved. A vision guided landing system is proposed with a cooperative target design based on infrared light as a landing target. It can recognize the cooperative target successfully in different light intensities and complex peripheral environments, and has good antinoise ability. A set of guidance system for image processing and position calculation is designed, and experiment results show that,the error of the attitude parameter is less than 2° while the error of the position parameter is less than 2cm, both meeting the precision requirements of landing, and the processing speed of each frame is about 30ms, which demonstrates excellent real-time performance.