CAJ | 학술논문

基于无人机影像的光束法空三是目前航空摄影测量领域的研究热点之一。本文简要分析了无人机航测与传统航测相比出现的技术难点；通过对无人机光束法空三平差中稀疏矩阵的分块存储和解算，节省了计算机内存的消耗，提高了运算速度，解决了无人机航测大数据量导致的问题；并通过对可能引起矩阵相关性的点的检测和剔除，有效避免了病态矩阵的出现，解决了无人机航测数据质量不好导致的问题。采用VS2010 C＋＋开发实现，通过对无人机实验数据的处理，表明该算法具有可行性与高效性，满足1∶2000的无人机测图要求。
기우무인궤영상적광속법공삼시목전항공섭영측량영역적연구열점지일。본문간요분석료무인궤항측여전통항측상비출현적기술난점；통과대무인궤광속법공삼평차중희소구진적분괴존저화해산，절성료계산궤내존적소모，제고료운산속도，해결료무인궤항측대수거량도치적문제；병통과대가능인기구진상관성적점적검측화척제，유효피면료병태구진적출현，해결료무인궤항측수거질량불호도치적문제。채용VS2010 C＋＋개발실현，통과대무인궤실험수거적처리，표명해산법구유가행성여고효성，만족1∶2000적무인궤측도요구。
Bundle aerotriangulation based on UAV images is a hot topic in the research of aerial photogrammetry . This article briefly analysed UAV aerial technical difficulties compared with traditional aerial .By decomposing and cal-culating sparse matrixes existing in bundle aerotriangulation adjustment ,much computer memory is saved ,which acceler-ates operation and solves problems caused by large amount of data in UAV aerial ;Ill-conditioned matrix is avoided through detecting and removing points that may cause matrix correlation ,which solves problems caused by bad data in UAV aerial.The adjustment method has been developed in C++.The result of experiment based on UAV images showed that the method is efficient and feasible .The adjusting results also may meet surveying and map in 1 ∶2000 .