Urban building automatic extraction method suitable for large-scale regional remote sensing image

Abstract

The invention discloses an urban building automatic extraction method suitable for a large-scale regional remote sensing image. The extraction method comprises the following steps: acquiring a high-resolution remote sensing image and an urban road network map; making a low dense building data set by using the remote sensing image and the road network map, and making a sparse single building data set by using the remote sensing image; training a low dense building data set by adopting a semantic segmentation model, and training a sparse single building data set by adopting an instance segmentation model; splicing results of the two models and then post-processing to obtain a final urban building distribution map. According to the invention, the problem of non-uniform distribution of buildings in mathematical significance and the phenomenon that the buildings are easily confused with other ground objects in the extraction process of the large-scale urban buildings are effectively solved, a high-precision urban building distribution diagram is realized, and the method has important significance for engineering application.

Description

technical field [0001] The invention relates to the field of image processing, in particular to a method for automatically extracting buildings at the urban scale level based on high-resolution remote sensing images, which can be widely used in high-resolution building distribution mapping, urban planning, path planning, and smart cities Practical applications. Background technique [0002] The high-resolution remote sensing images obtained by aviation or spaceflight are a means of obtaining space data. High-resolution remote sensing images play an important role in ground object recognition, time series analysis, insurance claims, emergency observation, disaster management, etc., greatly reducing the labor required by traditional methods and providing a new kind of data. Urban buildings are one of the important features of a city, and one of the main factors to measure the size of a city. It is very important in urban development planning, urban landscape design, disaster ...

AI Technical Summary

Problems solved by technology

There are two main reasons why the extraction of urban buildings is difficult to achieve at the application level: (1) The distribution of urban buildings is dense and sparse, and the buildings in urban villages and other areas are mainly dense and low buildings. The area is dominated by single-family high-rise buildings. The buildings in these different areas do not obey the same distribution in the mathematical sense, which restricts the extraction effect of urban buildings in large-scale areas, resulting in limited practical applications.

(2) The shape, size, color, and texture of urban buildings are rich in diversity, and they are complex in high-resolution remote sensing images. In the extraction results obtained by the model, buildings are easily confused with other ground objects, resulting in poor accuracy. Low, the effect is not good in practice

Images