CAJ | 학술논문

为了扩大视场范围，更好地利用无人机遥感图像和定位数据，针对现有的基于特征的图像拼接算法较慢以及无法提供地理信息的问题，实现了一种带地理信息的无人机遥感图像的快速拼接系统。考虑无人机在单航带内垂直下视的拍摄条件，对几何校正模型进行了化简，采用了尺度不变特征变换（ Scale Invariant Feature Transform， SFIT）方法配准图像序列，利用GDAL开源库和地理信息快速拼接航带内的遥感图像。实验表明，该系统能够满足实际工作的需要，为不同航带间和摆扫拍摄条件下的遥感图像拼接打下了基础。
위료확대시장범위，경호지이용무인궤요감도상화정위수거，침대현유적기우특정적도상병접산법교만이급무법제공지리신식적문제，실현료일충대지리신식적무인궤요감도상적쾌속병접계통。고필무인궤재단항대내수직하시적박섭조건，대궤하교정모형진행료화간，채용료척도불변특정변환（ Scale Invariant Feature Transform， SFIT）방법배준도상서렬，이용GDAL개원고화지리신식쾌속병접항대내적요감도상。실험표명，해계통능구만족실제공작적수요，위불동항대간화파소박섭조건하적요감도상병접타하료기출。
Current UAV remote sensing images mosaicking methods face serious challenges such as high computational cost and lack geographic information.We achieve a fast mosaicking system with geographic information in order to expand the field of view and make a better use of UAV images and positioning data.Considering the shooting conditions in single strip,we simplify the geometric cor-rection model. SIFT algorithm is used for registration image sequences. Using GADL database and the geographic information, we complete the fast mosaic of remote sensing images.Test results show that this system can satisfy actual requriement.Furthermore,it lays a foundation for images mosaic in sweeping shooting conditions and between different flight strips.