A port detection method on high-resolution remote sensing images

Abstract

A port detection method on high-resolution remote sensing images. The purpose of the invention is to solve the problems existing in the prior art. A non-supervised method is used to extract the separable patterns of ports in remote sensing images as middle-level features, and the extracted Middle-level features are used to detect ports of different scales and directions in high-resolution remote sensing images. The method of the invention has strong generalization ability for different types of ports in high-resolution remote sensing images, and can detect various special types of ports.

Description

technical field [0001] The invention belongs to remote sensing image (high spatial resolution, about one meter) processing technology. Background technique [0002] Considering that in order to berth ships, there are usually multiple concrete structures extending to the waters in the nearshore area, which are jagged at the junction of the waters. Li Yan et al., Fang Xiaofu et al. The structure performs port detection. However, only relying on the concave-convex structure is easy to misdetect the features with similar structures, such as peninsulas and river estuaries. Therefore, Liu Chun et al. proposed to add the length of the straight line segment, the angle between the inflection points, and the degree of aggregation of the shoreline corner points on the basis of detecting the concave-convex structure. and other constraints to improve the accuracy of detection [5] . Other scholars, such as Wu Jianhua, Wei Junwei, Liu Lili, and Chen Qi, have tested ports by detecting th...

AI Technical Summary

Problems solved by technology

These methods have poor generalization ability for different types of ports in high-resolution remote sensing images, and can only detect several special types of ports

Images