普通数码影像光线束算法特性分析

Analysis of the Bundle Adjustment Based on Common Digital Images

光束法是普通数码影像量测的最主要数据处理方法,有着精度高、易于引入系统误差改正等优点。但是,该算法也存在着对控制点数量要求多、未知数多且相互干扰大等缺点。这些缺点在不同的应用条件下,其影响会得以放大或缩小,然而至今对光束法的评价大都限于检查点位中误差,缺乏对各影响因素系统的、定量的研究。本文选择关键因素开展光束法算法试验及特性分析,明确算法特性,得出关键结论并对光线束算法的应用起到具体指导作用。

Bundle adjustment is the main common digital image data processing method of measurement, With high precision, easy to introduce systematic error correction, etc. However, the algorithm also exist on the number of control points required more and more unknowns and disadvantages such as mutual interference. These shortcomings in different application conditions, its impact will be able to zoom in or out. So far, however, most of the evaluation method is limited beam checkpoint bit errors, lack of quantitative research on the various influencing factors.This paper selected key factors to carried out bundle adjustmenthe test and analyzed algorithm characteristic.Explicit algorithm characteristics and play a specific role in guiding the application of bundle adjustment.