Cross-source remote sensing data target identification method based on symbol distance characteristics

Abstract

The invention discloses a cross-source remote sensing data target recognition method based on symbol distance characteristics, and belongs to the technical field of remote sensing image target recognition. In order to solve the problem of large difference between heterogeneous image feature extraction and ground feature recognition in remote sensing data target recognition, the method comprises the following steps: firstly, obtaining a to-be-recognized remote sensing image, and inputting the remote sensing image into a symbol distance feature extraction network for symbol distance feature extraction to obtain a distance value s from a random sampling point to a target boundary; when the input data is only homologous data, the boundary point set is taken as a final boundary point set, and when the input data is heterologous data of the same target, symbol distance features corresponding to the visible light image, the infrared image and the SAR image are fused according to a proportion, whether points belong to the surface of the target is judged, and the final boundary point set is obtained; and then performing three-dimensional feature extraction on the obtained three-dimensional point set to obtain a three-dimensional feature vector, and identifying the feature vector XE through a classifier. The method is mainly used for target identification of remote sensing data.

Description

technical field [0001] The invention relates to three-dimensional feature extraction and category recognition of ground objects in remote sensing images, and belongs to the technical field of remote sensing image target recognition. Background technique [0002] Remote sensing image target recognition technology has important research significance in both civilian and military aspects. Its purpose is to detect objects or areas of interest from static images, and to distinguish, detect and locate targets in the image. Accurate target recognition also lays the foundation for further completion of more complex tasks. Remote sensing image capture sensors mainly include: visible light, infrared and radar, etc. According to the similarities and differences of their sensor information sources, remote sensing images can also be divided into two categories: one is the data collected under the same information source; the other is the data collected from different sources. data. The...

AI Technical Summary

Problems solved by technology

[0004] In order to solve the problem that there are large differences between heterogeneous image feature extraction and ground object recognition in remote sensing data target recognition, the present invention proposes a cross-source remote sensing data target recognition method based on symbol distance features

Images