研究目的
To propose a primitive-based 3D building roof modeling method by integrating LiDAR data and aerial imagery, aiming to generate 3D building models with high accuracy in both the horizontal and vertical directions.
研究成果
The proposed primitive-based 3D building roof modeling method effectively integrates LiDAR data and aerial imagery to generate accurate building models. The method demonstrates higher geometric accuracy in both horizontal and vertical directions compared to traditional methods. Future work will focus on expanding the library of building primitives and improving the reconstruction of complex buildings.
研究不足
The proposed method is model-driven, which means differences between real buildings and the primitives can lead to a decrease in geometrical accuracy. Complex buildings with many dormers or those that do not fit predefined primitives may not be accurately reconstructed.
1：Experimental Design and Method Selection:
The study uses a primitive-based approach to represent building roofs and constructs a cost function using constraints from both LiDAR data and aerial imagery. The methodology involves optimizing the parameters of geometric primitives to fit the data from both sources.
2：Sample Selection and Data Sources:
The study uses both simulated data and real data provided by ISPRS, including LiDAR point clouds and aerial images of residential areas.
3：List of Experimental Equipment and Materials:
Airborne Laser Scanner (ALS) data from a Leica ALS50 system and digital CIR images from an Intergraph/ZI DMC (Digital Mapping Camera).
4：Experimental Procedures and Operational Workflow:
The process includes building detection and segmentation, feature extraction (rooftop planar patches and corners), and building modeling through primitive parameter optimization.
5：Data Analysis Methods:
The accuracy of the reconstructed building models is evaluated by comparing them to reference data, using root mean square error (RMSE) in both horizontal and vertical directions.
独家科研数据包，助您复现前沿成果，加速创新突破
获取完整内容
