论文部分内容阅读
基于七参数正形变换的数据驱动模型实现了机载LiDAR条带平差,算法借鉴了Robert(2004)的最小二乘表面匹配思想,通过引入高斯-马尔科夫模型改进了原有算法,得到未知参数的最小无偏方差估计。实验采用两组实测数据,分别考察了引入高斯-马尔科夫模型的必要性、算法效率以及迭代收敛性和算法精度。实验表明:(1)剖面检查吻合且精度一致;(2)Terra Match量测匹配精度,理想数据高程匹配误差小于0.05m,数据质量不理想时误差稍大,但均能成功匹配。
The data-driven model based on the seven-parameter orthomorphism transform realizes the on-board LiDAR strip adjustment. The algorithm draws on Robert (2004) least square surface matching idea, improves the original algorithm by introducing Gaussian-Markov model, and obtains The least unbiased variance estimation of unknown parameters. Two groups of experimental data are used in the experiment. The necessity of introducing Gaussian-Markov model, the efficiency of the algorithm, the iterative convergence and the accuracy of the algorithm are investigated respectively. Experiments show that: (1) the cross-sectional inspection is consistent and the precision is consistent; (2) Terra Match measurement matching accuracy, the error of the ideal data elevation matching is less than 0.05m, the error is slightly larger when the data quality is not ideal, but all can be successfully matched.