A method of recovering distorted DEM of regular terrain????(Open Access)

作者：Xu, Junbo; Zhou, Bo^*

来源：Revista Tecnica de la Facultad de Ingenieria Universidad del Zulia, 2016, 39(10): 264-270.

DOI：10.21311/001.39.10.31

• Digital elevation model
• Equation of plane
• Least squares methods
• Parameter estimation algorithm
• Random sample consensus
• Random sample consensus (RANSAC) algorithm
• Recovery methods
• Terrain features

The distortion often occurs in digital elevation model (DEM) depicting terrain, and there is no good strategy to recover distorted DEM. It is necessary to make a recovery for a high-accuracy and high-fidelity DEM. In this paper, a method is proposed to recover the DEM which is used to depict regular terrain by taking a flat terrain object as an example. According to the terrain feature of flat terrain, it should be a plane of DEM, and the equation of plane can be solved by parameter estimation algorithms. This research selects the Least-Squares method (LS) and the Random Sample Consensus algorithm (RANSAC) to estimate the parameters of equation. Based on the plane equation, the recovered DEM can be obtained. The result shows that the recovery method is effective and feasible, and the recovered DEM not only has a higher accuracy, but can reflect terrain feature fully and exactly. With a large quantity of grid points, the recovering capacities of two algorithms are similar. However, compared with the LS, the RANSAC can adapt to the mutations of elevation with a better robustness.

• 出版日期2016
• 单位合肥工业大学