An automatic geometric correction method for remote sensing image data based on deep learning

Abstract

The invention discloses an automatic geometric correction method for remote sensing image data based on deep learning, comprising: step 1, selecting road intersections as control points from vector road data according to the geometric topological relationship; step 2, using vector data road The intersection control point is the center, and the remote sensing image is intercepted with a certain area of ​​the window, and the image road intersection is extracted from the intercepted remote sensing image by using the trained deep learning model, and the control point of the same name is formed by matching the geometric distance with the vector data road intersection; Step 3: Use the density-based spatial clustering algorithm to clean the data of the control points with the same name, and use the cleaned control points with the same name to perform geometric correction on the remote sensing image based on the binary cubic polynomial correction model. In the present invention, a deep learning model is used to automatically extract road intersection control points from remote sensing images, which improves the accuracy of feature extraction from remote sensing images.

Description

technical field [0001] The invention relates to the technical field of geometric correction of remote sensing image data, in particular to an automatic geometric correction method of remote sensing image data based on deep learning. Background technique [0002] At present, the spatial resolution of multi-band remote sensing images has reached the meter level, and the spatial resolution of single-band images has reached sub-meters. Such high-resolution remote sensing images can clearly express the structure, texture and other detailed information of some ground features and landscapes, so that we can not only obtain rich spectral information of ground features, but also obtain more information about the structure, shape and texture of ground features. The information makes it possible to observe the detailed changes of the earth's surface on a small spatial scale, to carry out large-scale remote sensing mapping, and to monitor the impact of human activities on the environmen...

AI Technical Summary

Problems solved by technology

[0006] The present invention realizes automatic geometric correction of remote sensing images by providing an automatic geometric correction method of remote sensing image data based on deep learning, and solves the problem of low efficiency of artificial remote sensing image geometric correction

Images