CAJ | 학술논문

研究云计算环境下无人机采集无序图像快速拼接问题。无人机采集图像中，飞行速度、位置都呈现不可控性，采集的大量图像也呈现无序状态，存在重复采集、缺失采集的情况，在传统的云计算环境下，拼接图像需要大量重复认证、对比，匹配过程效率很低。为此提出了一种基于哈密顿通路模型的云计算下无人机采集无序图像快速拼接方法。对无人机采集的无序图像进行空间位置几何畸变补偿，建立哈密顿通路模型，将云计算下无人机采集无序图像拼接问题转换为哈密顿通路模型求解问题，利用哈密顿通路求解的快速性实现云计算下无人机采集无序图像快速拼接方法。实验结果表明，利用改进算法进行云计算下无人机采集无序图像快速拼接，能够提高图像的匹配效率，缩短拼接时间。
연구운계산배경하무인궤채집무서도상쾌속병접문제。무인궤채집도상중，비행속도、위치도정현불가공성，채집적대량도상야정현무서상태，존재중복채집、결실채집적정황，재전통적운계산배경하，병접도상수요대량중복인증、대비，필배과정효솔흔저。위차제출료일충기우합밀돈통로모형적운계산하무인궤채집무서도상쾌속병접방법。대무인궤채집적무서도상진행공간위치궤하기변보상，건립합밀돈통로모형，장운계산하무인궤채집무서도상병접문제전환위합밀돈통로모형구해문제，이용합밀돈통로구해적쾌속성실현운계산하무인궤채집무서도상쾌속병접방법。실험결과표명，이용개진산법진행운계산하무인궤채집무서도상쾌속병접，능구제고도상적필배효솔，축단병접시간。
In this paper, the problem of quickly stitching of disorderly images acquired by unmanned aerial vehi-cle ( UAV) under cloud computing environment was researched. W proposed a method for quickly stitching of disor-derly images which are acquired by unmanned aerial vehicle ( UAV) under cloud computing environment based on the Hamiltonian path model. Firstly, the disorderly images acquired by UAV were conducted to make geometric distortion compensation of spatial position and construct the Hamiltonian path model. Then, the disorderly image stitching prob-lem was converted into the problem of Hamiltonian path model under the cloud computing environment. Meanwhile, by taking the advantage of the fast solving features of Hamiltonian path, the quickly stitching of disorderly images ac-quired by UAV was ultimately accomplished under cloud computing environment. Experimental results show that the improved algorithm for quickly stitching of disorderly images acquired by UAV under cloud computing environment can enhance the efficiency of image matching and shorten stitching time.