空地协同视觉实现无人机自主降落研究

Research on UAV Autonomous Landing with Air-ground Collaborative Vision

在深入分析了无人机视觉降落多种方式的基础上,结合无人机连续视觉图像的三维空间信息,提出了基于空地协同视觉信息的无人机自主降落方法,设计了一种多图形组合、多色彩导引、多方法识别的地面合作目标。从坐标系的转换、合作目标的选取、识别与处理等方面进一步细化论述空地协同视觉完成无人机自主降落的方式,通过灰度变换、HSV色彩变换、HU不变距匹配等方法保障了合作目标的准确识别率。

Based on the in-depth analysis of various modes of UAV(Unmanned Aerial Vehicle) visual landing combined with three-dimensional spatial information of UAV continuous visual imagery,we propose a UAV autonomous landing method based on the cooperative visual information of open space and design a multi-graphic combination with multi-color guidance and multi-method identification of ground cooperation goals.From the transformation of the coordinate system,and the selection,recognition,and processing of the cooperation target,this paper further elaborates the application of air-ground cooperative vision to complete drone autonomous landing,and guarantees methods of grayscale transformation,HSV color transformation,and HU constant distance matching with accurate recognition rates of cooperation goals.