Remote sensing image oil tank detection method integrated with semantic model

Abstract

The invention discloses a remote sensing image oil tank detection method integrated with a semantic model. The method comprises the following steps: acquiring data: acquiring remote sensing image sample data and preprocessing the remote sensing image sample data; constructing models: constructing an oil tank detection model and a semantic model; model training; inputting the processed image into the oil tank detection model for training to obtain a feature map and feature map detection information, establishing a relationship of the preprocessed image classification information, storing and establishing a scene knowledge base, and inputting an information relationship of the scene knowledge base into the semantic model for training to obtain image scene context information; and model fusion: fusing the detection information of the feature map with the image scene context information. By the adoption of the method, oil tank detection is conducted according to different image scales, image information is enhanced, and better oil tank detection precision is achieved.

Description

【Technical field】 [0001] The invention belongs to the technical field of image processing, in particular to a remote sensing image oil tank detection method integrated into a semantic model. 【Background technique】 [0002] With the continuous development of artificial intelligence, object detection has attracted extensive attention from academia and industry. Oil tanks are an important tool for resource storage, and the detection of oil tanks in remote sensing images is of great help to military affairs and geography. The traditional oil tank target detection method is usually a candidate target area. The training model is used to extract features and classify objects into different categories, and the bounding box of the classified area is obtained for detection. It is affected by the shooting angle, weather conditions and target density. There can be a general method to detect oil tanks in remote sensing images fused with image contextual information. [0003] At present...

AI Technical Summary

Problems solved by technology

[0004] The present invention provides a remote sensing image oil tank detection method integrated into the semantic model. The integration method uses deep learning to improve the accuracy of oil tank detection and overcomes the problem that traditional oil tank detection cannot be integrated with image context information

Images