基于深度学习的无人机航测三维重建算法优化  

Optimization of 3D reconstruction algorithm for UAV aerial survey based on deep learning

作　　者：杨永刚[1] 李思萌 YANG Yonggang;LI Simeng(College of Transportation Science and Engineering,CAUC,Tianjin 300300,China;Intelligent Network Institute,China Information Security Research Institute Co.,Ltd.,Beijing 102200,China)

机构地区：[1]中国民航大学交通科学与工程学院,天津300300 [2]中国信息安全研究院有限公司智能网络所,北京102200

出　　处：《中国民航大学学报》2024年第6期82-90,共9页Journal of Civil Aviation University of China

摘　　要：利用传统的运动恢复结构(SFM,struct from motion)算法实现无人机视角下的三维重建时,为减少特征点的错误匹配和运动目标对整体稀疏点云的影响,主要依靠随机采样一致性(Ransac,random sample consensus)算法,但会导致Ransac在求解相机位姿时的正确率降低,迭代次数增加。本文基于深度学习的单次多框检测器(SSD,single shot multibox detector)网络进行目标检测。首先,在尺度不变特征转换(SIFT,scale-invariant feature transform)提取特征点后,对动态目标类别范围内的特征点进行剔除;然后,在K近邻(KNN,K-nearest neighbor)暴力匹配后,对错误匹配进行剔除,减少无效运动目标范围内的特征点和不同类别间的错误匹配,使得相同置信度时,Ransac求解相机位姿的迭代次数降低,同时减少特征点暴力匹配和SFM算法计算三维点的时间。最后,通过2个场景12张图片验证了引入深度学习优化后三维重建算法的可行性。When the traditional struct from motion(SFM)algorithm is used to realize 3D reconstruction from the perspective of unmanned aerial vehicle(UAV),in order to reduce the mismatching of feature points and the impact of moving targets on the overall sparse point cloud,the random sample consensus(Ransac)algorithm is mainly relied on.However,these problems can lead to a decrease in the accuracy and an increase in the number of iterations of Ransac when solving camera poses.This article conducts target detection based on a deep learning single shot multibox detector(SSD)network.Firstly,feature points within the range of dynamic target categories are removed after scale-invariant feature transform(SIFT)extraction of feature points.Then,mismatches are removed after K-nearest nei-ghbor(KNN)violent matching to reduce feature points within the range of invalid moving targets and mismatching between different categories.So that when the confidence is the same,the number of iterations of Ransac when solving camera pose is reduced,and the time of feature point violence matching and SFM algorithm calculation of 3D points are also reduced.Finally,the feasibility of the 3D reconstruction algorithm optimized by deep learning was verified through 12 images of 2 scenes.

关 键 词：无人机(UAV) 三维重建 目标检测 单次多框检测器(SSD) 

分 类 号：V279.2[航空宇航科学与技术—飞行器设计]