飞行昆虫视觉导航的仿生算法
首发时间:2014-12-30
摘要: 果蝇,蜜蜂等飞行昆虫具有简单小巧的大脑结构,这种能帮助它们完成稳定着陆,导航至食物源,障碍物回避,飞行姿态控制等操作的精巧大脑结构,吸引了越来越多的研究者对昆虫大脑的飞行机理进行研究。许多昆虫专家的实验表明,果蝇,蜜蜂等昆虫通过飞行过程中接收到的图像流动信息来确定食物位置以及各种导航任务。研究昆虫导航机理有助于开发新的视觉导航算法,对仿生类导航算法有着重要意义。本文总结了昆虫导航机理,并且提出了一种简单的熵流概念用来仿生昆虫视觉导航。并且提出了创新的熵流自动评估阈值选取算法来自适应的剔除干扰数据。仿真实验表明,利用图像纹理来进行的导航能够有效的引导飞行器飞至目标位置。
For information in English, please click here
The bionic navigation algorithm for the flying insect
Abstract:Flying insects like fruit fly and bee have simple but delicate brain structure, which make them handle the flying mission like smooth landing, navigation to the food source and obstacle avoiding successfully. These attracted tremendous researchers to study the visual mechanism of insects. Many experts present through the experiment that flying insects like bee and fruit fly navigate themselves to the food source by the motion of the visual images. There will be great significance for innovated navigation system by studying the visual mechanism of bee. This paper give a brief introduction for the vision mechanism of flying insects, and present a simple concept of entropy flow for bionic-navigation system, and present a self-adaptation election algorithm to get rid of noisy data. Simulation results show that the navigation using the texture of images can be efficient and effective method to navigate the drone to the correct destination.
Keywords: Navigation guidance and control bionic navigation entropy flow self-adaptation election algorithm
论文图表:
引用
No.4623631102553014****
同行评议
共计0人参与
勘误表
飞行昆虫视觉导航的仿生算法
评论
全部评论0/1000