Method for automatically annotating remote sensing images on basis of deep learning

Abstract

The invention discloses a method for automatically annotating remote sensing images on the basis of deep learning. The method for automatically annotating the remote sensing images includes extracting visual feature vectors of the to-be-annotated remote sensing images; inputting the visual feature vectors into a DBM (deep Boltzmann machine) model to automatically annotate the to-be-annotated remote sensing images. The DBM model implemented in the method sequentially comprises a visible layer, a first hidden layer, a second hidden layer and a tag layer from bottom to top, and is acquired by means of training. The method for automatically annotating the remote sensing images has the advantages that the deep Boltzmann machine model implemented in the method comprises the two hidden layers (namely, the first hidden layer and the second hidden layer respectively), accordingly, the problem of 'semantic gaps' in image semantic annotation procedures can be effectively solved by the two hidden layers, and the integral annotation accuracy can be improved.

Description

technical field [0001] The invention relates to intelligent classification and retrieval technology of remote sensing images, in particular to an automatic tagging method for remote sensing images based on deep learning. Background technique [0002] Remote sensing images are one of the important data of spatial information, and are widely used in geological and flood disaster monitoring, agricultural and forest resources investigation, land use and urban planning, and military fields. With the development of my country's space science and earth observation technology, the data of remote sensing image data shows an exponential growth trend every year, and the effective management of massive remote sensing image data has become increasingly important. [0003] Remote sensing image annotation is one of the important contents of remote sensing image analysis and understanding. It extracts the underlying visual features of remote sensing images and learns the connection between ...

AI Technical Summary

Problems solved by technology

Most of the traditional image annotation work is to analyze and understand the visual content of the image through the low-level visual features of the image, but most of this method has a problem: "semantic gap"

"Semantic gap" means that the high-level semantics of the image cannot be inferred only through the low-level visual features of the image. There is no suitable abstraction bridge between the low-level visual features of the image and the high-level semantics of the image, so the labeling effect is not ideal.

[0006] In order to overcome the problem of "semantic gap", people have gradually developed some methods to map the low-level visual features of the image to the high-level semantics of the image. Typical methods include the Probabilistic Latent Semantic Analysis (pLSA) model, latent Dirichlet Allocation (Latent Dirichlet Allocation, LDA) model and Author Topic Model (ATM) model, etc., but most of these methods only consider the color texture characteristics of the image, and do not consider the spectral characteristics of remote sensing images