基于帧差法改进的无人机运动检测研究  

Research on an Improved Method for the Multi-UAV Motion Detection Based on Inter-frame Difference

作　　者：王茂森[1] 陈龙[1] 戴劲松[1] 

机构地区：[1]南京理工大学机械工程学院,江苏南京210094

出　　处：《机械制造与自动化》2016年第3期165-168,共4页Machine Building & Automation

摘　　要：在无人机群组飞行过程中,基于主动视觉的算法是目前用于编队与协同飞行测控的主要方法。研究了一种改进算法,先改进了帧间差分法使其能够更加准确地获得无人机的位置信息,然后在前者检测到的运动区域中使用水平集方法提取运动轮廓,从而提高对无人机运动检测的准确性,也为后续的目标识别做好了基础。仿真实验验证了该方法的可行性。For UAV flight, the active vision algorithm is the main method to detect their formation and cooperative flight at present. But research shows that modified inter-frame difference can obtain UAV location information move accurately. Then in the area of movement, their motion profile can be detected by level set method, which can further improve the accuracy of motion detection, and lay foundation for subsequent target identification. This simulation experiment verifies the feasibility of the new method.

关 键 词：无人机 运动检测 帧间差分法 形态学图像处理 水平集法 

分 类 号：V279[航空宇航科学与技术—飞行器设计]