Wide Area Situation Recognition by Portable Aerial Image Analyzing System

从2009年到2012年，我们实验制作了便携式航空图像采集系统（广告气球式），并开发了静态和动态环境下的图像恢复算法。在室内人体行走场景和室外汽车运动场景中，验证了便携式航空图像采集系统和图像恢复算法的有效性。自2013年至2014年期间，利用便携式直升机（AR无人机）积极获取广域图像。我们新开发了AR地标识别方法和基于航拍图像的人体跟踪方法。室内体育馆和室外大学道路的实验结果表明了该方法的有效性。

Since 2009 till 2012 year period, we experimentally produced a portable aerial image acquisition system (ad balloon type) and developed the image restoration algorithm in static and dynamic environment. Both the effectiveness of the portable aerial image acquisition system and the image restoration algorithm had been proved in indoor human walking scenes and outdoor car moving scenes.Since 2013 till 2014 year period, a portable helicopter (AR Drone) is utilized to actively acquire wide area images. We newly developed the AR landmark recognition method and aerial image based human tracking method. Experimental results both in an indoor gymnasium and outdoor university roads have shown the effectiveness of the proposed methods.