全文总字数：14202字
基于视觉的无人机降落的障碍识别的文献综述
摘要
本设计在无人机上装双目摄像头，拍摄出降落区域的照片，对图像进行处理，进行特征点的提取，再对提取的特征点进行匹配，优化匹配，再放大特征点倍数，设立特征点集合，最后对障碍物进行区域构建，组建障碍物的信息。视觉技术导航方式不受障碍物外观轮廓、外界信号干扰的影响，测量范围广泛，随着计算机硬件技术以及智能算法的快速发展,利用计算机视觉进行避障的反应速度也越来越快。视觉设备重量轻、体积小、价格低廉适合民用旋翼无人机载重较轻的特点。
关键词：双目视觉无人机，避障系统，障碍识别
Abstract
In this design, a binocular camera is installed on the UAV to take photos of the landing area, process the images, extract the feature points, match the extracted feature points, optimize the matching, enlarge the multiple of the feature points, set up the feature point set, and finally construct the area of the obstacles to form the information of the obstacles. With the rapid development of computer hardware technology and intelligent algorithm, the response speed of using computer vision to avoid obstacles is faster and faster. The vision equipment is light in weight, small in size and low in price, which is suitable for the civil rotorcraft with light load.
Key words: binocular vision UAV, obstacle avoidance system, obstacle recognition
1、研究背景
以上是毕业论文文献综述，课题毕业论文、任务书、外文翻译、程序设计、图纸设计等资料可联系客服协助查找。
