Building extraction optimization method based on multi-temporal remote sensing image

Abstract

The invention discloses a building extraction optimization method based on a multi-temporal remote sensing image, and belongs to the technical field of satellite remote sensing image processing and application. The method comprises the following steps of: firstly, performing initial dichotomy of an optimized precision ratio on a multi-temporal remote sensing image, realizing a binary classification process of buildings and non-buildings, and reducing misclassification in the dichotomy process by adopting texture, geometry and area features; secondly, carrying out object intersection and union operation on the initial binary extraction results of the multiple time phases, carrying out union operation on intersection results, and traversing all time phase extraction results; and finally, optimizing the non-building class by using the neighborhood relative boundary index, and reducing cavity removal and edge vacancy in the building class. Experimental results of three-time-phase high-resolution remote sensing images show that the building recognition rate can be improved, noise interference of non-buildings can be reduced, a good extraction result is achieved, and compared with single recognition, the extraction precision is obviously improved.

Description

technical field [0001] The invention belongs to the technical field of satellite remote sensing image processing and application, and in particular relates to a building extraction optimization method based on multi-temporal remote sensing images. Background technique [0002] The use of remote sensing images for automatic and semi-automatic building extraction is of great significance. On the one hand, in the face of massively increased remote sensing image update data, building categories can be updated more quickly and efficiently, thereby contributing to the realization of digital earth and smart cities. Requirements for visualization and statistics of building categories. On the other hand, it can reduce the traditional manual visual interpretation work mode and reduce manpower and labor costs. [0003] However, existing methods still have many difficulties for building extraction. One of the notable difficulties is the occlusion of shadows of trees and tall buildings...

AI Technical Summary

Problems solved by technology

[0003] However, there are still many difficulties in the existing methods for building extraction

One of the notable difficulties is the occlusion of trees and tall building shadows, which makes the results obtained by using the spectral difference of building roofs to be extracted incomplete and easily confused with surface shadows

Among the existing methods, the traditional supervised classification and shadow migration algorithms are more based on a single image to improve the accuracy. Without introducing more data sources, the upper limit of the accuracy of single image extraction is limited.

Images