跑道
计算机科学
航空电子设备
计算机视觉
过程(计算)
地形
人工智能
航空影像
匹配(统计)
实时计算
遥感
图像(数学)
工程类
地理
统计
地图学
数学
考古
航空航天工程
操作系统
作者
Jingzhe Wang,Yue Cheng,Jianchun Xie,Wensheng Niu
摘要
With the development of avionics technology, the runway detection technique becomes increasingly important both in military and civil applications. But the existing runway detection methods are mostly intended for satellite images or downward-looking aerial photos and the algorithm complexity is relatively high, which makes it hard to apply them to airborne real-time applications. In this paper, a real-time sensor guided runway detection method is presented. First of all, with the help of sensor data and terrain data from EVS and SVS, a runway template is generated and a search region is therefore determined. Then within the search region, a lines extraction on query image is performed. And finally the matching process is applied to the query image and template to detect the runway area. Experimental results on simulation system and real aerial images show that the proposed approach has a high precision and an 87.7% less time cost than an existing state-of-art matching method.
科研通智能强力驱动
Strongly Powered by AbleSci AI