Photogrammetry point cloud filtering method fusing image information

Abstract

The invention discloses a photogrammetric point cloud filtering method for fusing image information. The method comprises the following steps: fusing images and photogrammetric point cloud elevation multi-source information; introducing a novel classifier to classify the fused images, and guiding the classification result as knowledge to be used in point cloud filtering, i.e., superposing a classification thematic map and photographic point cloud to filter non-ground points, realizing point cloud filtering and generating a digital elevation model. According to the point cloud filtering method,the problems that a conventional point cloud filtering algorithm is poor in universality, the filtering effect is limited by parameter setting, and image information for generating point cloud is notutilized are solved, the DEM generated through the point cloud filtering method is highly related to the reference DEM, and the workload of manual post-processing for producing the DEM can be greatlyreduced.

Description

technical field [0001] The invention relates to a photogrammetry point cloud filtering method, in particular to a photogrammetry point cloud filtering method for fusing image information. Background technique [0002] As one of the 4D products, digital elevation model (digital elevation model, DEM) has a wide range of applications in the fields of surveying and mapping, land and national defense construction. At present, surveying and mapping satellites or aerial images are used as data sources in land survey and mapping, and photogrammetry point clouds are obtained by using image matching technology, and DEM is generated after post-processing. However, in the production process with a high degree of automation in mapping, the automatic filtering technology of photogrammetric point cloud is still immature, and the post-processing of DEM requires a lot of manpower and material resources. important meaning. [0003] Most of the traditional point cloud filtering algorithms ar...

AI Technical Summary

Problems solved by technology

However, in the production process with a high degree of automation in mapping, the automatic filtering technology of photogrammetric point cloud is still immature, and the post-processing of DEM requires a lot of manpower and material resources. important meaning

[0003] Most of the traditional point cloud filtering algorithms are designed for LiDAR point clouds. The proposed LiDAR point cloud filtering algorithms mainly include slope-based filtering algorithms, mathematical morphology-based methods, and surface-based filtering algorithms. The above-mentioned point cloud filtering algorithms

Images