Abstract:

A novel method for digital surface model (DSM) generation based on triangulation irregular network （TIN） is proposed in order to increase the efficiency. Firstly, by establishing the virtual grid and interpolation, an initial triangulation network is quickly generated in linear time. Secondly, the optimal TIN will be finally acquired by optimizing the local data in the initial triangulation network. Finally, the high precision DSM is obtained by processing the redundant and intensive data.The experiment results not only indicate that the triangulation efficiency of the proposed algorithm is much higher than those algorithms available, but also prove the effectiveness of the proposed algorithm for DSM generation when facing large amount of data.