Automatic acquisition method of deep learning sample library corresponding to remote sensing image land type recognition

Abstract

The invention belongs to the technical field of land use remote sensing monitoring, and particularly relates to an automatic acquisition method of a deep learning sample library corresponding to land type recognition in remote sensing images. Mark the point with a small gradient value in the remote sensing image as a mark point; perform flood filling through the mark point, and assign a mask corresponding to each filled area and save the terrain type information; extract the segmented image according to the mask, and According to the land type information of the land use status saved by the mask, it is classified and saved to form a sample library; the present invention realizes the automatic collection of remote sensing image feature libraries corresponding to different land types by superimposing and comparing the land use status data and remote sensing data in the same phase. Compared with the disadvantages of traditional manual sample acquisition such as heavy workload and difficult sample area acquisition, the sample acquisition method used in the present invention is faster and more accurate, and the labor cost is significantly reduced.

Description

technical field [0001] The invention belongs to the technical field of land use remote sensing monitoring, and in particular relates to an automatic acquisition method of a deep learning sample library corresponding to land type recognition of remote sensing images. Background technique [0002] In the field of land use status investigation technology, time-sensitive land use information is very important, and the automatic interpretation of land types from remote sensing images is a major technical problem that my country's land and resources science and technology are committed to solving! In recent years, with the rapid development of machine learning technology represented by deep learning, applying deep learning to automatic interpretation of remote sensing images and realizing land use type identification as automatic as possible is currently an important research goal and direction of Chinese researchers. However, the premise of deep learning corresponding to the work...

AI Technical Summary

Problems solved by technology

Traditionally, training images are mostly obtained manually and marked manually, which is time-consuming and labor-intensive, with a huge workload, and is easily affected by the operator's work emotions and work negligence

Images