应用神经网络实现基于等高线的三维地形图

Constructing Three-Dimensional Topographical Map Based on Contours by Applying Neural Network

在分析过程中,需要将其转换为三维地形图以利于更直观的观察,为此探讨应用神经网络方法实现基于等高线的三维地形图,先从等高线图中提取样本数据,进而利用神经网络的良好逼近功能,较好地拟合样本数据,从而可以得到其余部分的高程值。使用Mat-Lab的神经网络工具箱可以很方便地进行BP网络的建立、训练和仿真,并绘制出相应的三维地形图。实验表明,该方法的效果很好。

In the analysis process, it needs to be converted into three-dimensional topographical map to facilitate more intuitive observation. Explores constructing three-dimensional topographical map based on contours by applying neural network, after extracting the sample data from the contour map, one can fit the sample data better using the good approximation of the neural networks, thus gets the high value of the rest part of the map. The neural networks MatLab toolbox can be easily used to create, train and simulate a BP network and draws the corresponding three-dimensional topographical map. Experimental results show this method is effective.