Semantic Segmentation Method of Remote Sensing Image Based on Self-Supervised Contrastive Learning

Abstract

The invention discloses a remote sensing image semantic segmentation method based on self-supervised contrastive learning, comprising the following steps: constructing a semantic segmentation network model (such as Deeplab v3+); using unlabeled data to pre-train the encoder of the network model; pre-training is completed Finally, supervised semantic segmentation training is performed on the network model on the labeled samples; the remote sensing image is semantically segmented by using the network model completed by supervised semantic segmentation training; in the pre-training process, global style comparison and local matching comparison are used Combining methods for comparative study. The present invention applies comparative self-supervised learning to remote sensing semantic segmentation datasets, proposes a global style and local matching comparative learning framework, and forms a remote sensing image semantic segmentation method based on self-supervised comparative learning, making the semantic segmentation method more widely applicable , the segmentation effect is better.

Description

technical field [0001] The invention relates to the technical field of remote sensing image semantic segmentation, in particular to a remote sensing image semantic segmentation method based on self-supervised contrastive learning. Background technique [0002] With the development of remote sensing technology, it is easier to obtain high-resolution remote sensing images, and remote sensing images have more and more extensive applications in urban planning, disaster monitoring, environmental protection, transportation and tourism. The extraction and recognition of information in remote sensing images is usually the basis of all applications, and semantic segmentation is the technology of identifying and classifying full-image pixels, so it has always been an important and challenging research direction in the field of remote sensing. [0003] In recent years, with the development of deep learning technology, the semantic segmentation of remote sensing images has achieved impr...

AI Technical Summary

Problems solved by technology

[0008] In view of this, the purpose of the present invention is to solve the problem of directly learning features from unlabeled images to help downstream semantic segmentation tasks with only a small amount of annotations. In this paper, the contrastive self-supervised learning is applied to the remote sensing semantic segmentation data set, a global style and local matching contrastive learning framework is proposed, and a remote sensing image semantic segmentation method based on self-supervised contrastive learning is formed.

Images