基于神经网络的机载LIDAR高程数据系统误差纠正方法研究

Study on Error Correction Method of Airborne LIDAR Altitude Data System Based on Neural Network

该文介绍了BP神经网络算法的基本原理,引入了神经网络算法模型,通过神经网络算法构建控制点纠正模型来纠正实验区LIDAR测高数据并验证其精度。经验证,纠正后的实验区LIDAR测高数据精度和优良率都有所提高,粗差稳定,能够满足1∶10000DLG高程精度的要求,这样的LIDAR数据能够在基础测绘中得到应用和推广。

This paper introduces the basic theory of BP neural network algorithm,and it also introduces the neural network algorithm model. The control point correction model is built by neural network algorithm to correct the LIDAR altitude data in the experimental area and then the accuracy is verified.Through the verification,the accuracy and the excellent and good rate of the LIDAR altitude data in experiment area after the correction are all improved. The gross error is stable,it can meet the 1： 10000 DLG altitude accuracy requirement. This LIDAR data can be applicated and extended in basic surveying and mapping.