A Method for Intelligent Understanding of Remote Sensing Image Using Semantics to Describe the Spatial Relationship of Objects

Abstract

The invention provides a method for intelligent understanding of remote sensing images that describes the semantics of the spatial relationship of ground objects, including constructing a remote sensing image semantic understanding reference library and performing semantic intelligent description of remote sensing images. Building a remote sensing image semantic understanding reference library includes obtaining remote sensing image blocks, remote sensing The vector data block corresponding to the image block, the description sentence corresponding to the remote sensing image block, the target in the remote sensing image block and the relationship diagram between each target, etc.; the semantic intelligent description of the remote sensing image includes training and testing the model And intelligent description of remote sensing images based on the spatial relationship of ground objects for remote sensing image blocks outside the database. Applying the technical solution of the present invention, the effect is: realize the understanding of remote sensing images from different angles; the established high-level semantic reference library of remote sensing images allows adding description sentences corresponding to remote sensing images later, which is conducive to improving computer image understanding and description The ability; the processing method is streamlined and easy to implement.

Description

technical field [0001] The invention relates to the technical field of intelligent semantic understanding of image vision, in particular to a method for intelligent understanding of remote sensing images describing the semantics of the spatial relationship of ground objects. Background technique [0002] Remote sensing images are an important data source for geographic information research. It is a key and urgent task to continuously obtain rich visual information and deep hidden information from remote sensing images to achieve the purpose of understanding remote sensing images. Traditional remote sensing image understanding methods are based on image processing and image analysis, starting from the surface of the image, performing feature and statistical analysis, and analyzing and processing a single feature in the image. The result is also a simple mark, recognition and detection of the feature . It cannot achieve the scene-level understanding of the content in the larg...

AI Technical Summary

Problems solved by technology

However, these databases are only applicable to the field of natural images, and do not involve remote sensing images. For remote sensing images, which have rich sources, different resolutions, numerous image contents, complex ground objects, and many factors such as being easily affected by the external environment, it is difficult to achieve in-depth level of understanding

Images