Digital Terrain Model Derivation from Noisy Surface Data
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for converting Digital Surface Models (DSMs) into Digital Terrain Models (DTMs) are inefficient, especially when DSMs are noisy and lack additional information sources, requiring expensive and complex acquisition techniques like LIDAR, and often necessitate manual intervention.
Innovation Solution
A method that computes candidate surfaces to represent the ground surface using a random sample consensus method and an ad-hoc function derived from a known relation between DSM and DTM, allowing for automatic selection and merging of surfaces to form a single DTM, even in areas with significant noise and without additional information sources.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If LIDAR techniques are used to acquire high-quality DSMs, then measurement precision is improved, but cost and device complexity increase significantly
Solution Approach 1:
The patent replaces expensive LIDAR acquisition systems with cheaper photogrammetry-based DSM acquisition methods. By using multiple photographs taken from different positions and processing them through stereoscopic vision algorithms, the system achieves acceptable DSM quality without requiring sophisticated and costly LIDAR equipment.
Solution Approach 2:
The patent substitutes the mechanical LIDAR ranging system with an optical-photogrammetric system. Instead of using laser pulses to measure distances directly, the system uses camera photographs and computational stereoscopy to derive elevation information, replacing complex mechanical measurement systems with simpler optical capture and computational processing.
2Measurement precision
If manual intervention is used to identify ground surface points in noisy DSMs, then DTM accuracy is improved, but productivity decreases
Solution Approach 1:
The patent implements an automatic DTM extraction system that processes noisy DSMs without requiring manual intervention. The system uses clustering algorithms and surface fitting methods to automatically identify ground surface points and construct the terrain model, enabling the process to serve itself rather than requiring human operators.
Solution Approach 2:
The patent extracts the ground surface information from noisy DSM data by separating ground points from non-ground features using automated clustering and classification algorithms. This extraction process removes the need for manual point identification while maintaining accuracy through computational methods that isolate ground surface characteristics.
3Measurement precision
If additional information sources like cadastral maps are used to convert DSM to DTM, then DTM accuracy is improved, but device complexity and data processing requirements increase
Solution Approach 1:
The patent extracts ground surface information directly from the DSM data itself without requiring external information sources like cadastral maps or multi-spectra photography. By using clustering algorithms and surface analysis methods that work directly on the DSM elevation values, the system obtains DTM accuracy sufficient for radio coverage planning while avoiding the complexity of integrating multiple external data sources.
Data Source
Figure 1
Figure 2~13
Figure 3
AI summary
A method of deriving a digital terrain model from a digital surface model of an area of interest comprises: dividing the area of interest into a plurality of area portions or patches (205); calculating, from the digital surface model, a set of candidate surfaces adapted to represent a ground surface in each area portion (210); if such set includes at least two candidate surfaces, estimating a distance from the ground surface of each candidate surface by using a function of a set of geometrical features related to the considered candidate surface, such function being derived from a known relation between a digital surface model and the height of the ground surface in a reference area (215); selecting, as a representation of the ground surface in each area portion, the candidate surface having the smallest distance from the ground surface, so as to obtain local digital terrain models (220); and merging the different digital terrain models (225).