Remote sensing image intelligent interpretation method and system based on adaptive perceptron

Abstract

The invention discloses a remote sensing image intelligent interpretation method and system based on an adaptive perceptron, and relates to the technical field of remote sensing image target interpretation in deep learning, and the method comprises the steps: carrying out the preprocessing of a remote sensing image during the intelligent interpretation of the remote sensing image; constructing a self-adaptive perceptron based on a local threshold value; determining a threshold range and a perceptron weight; determining an interpretation network structure according to the adaptive perceptron and the improved UNet model; according to the network structure and the improved UNet remote sensing intelligent interpretation model, target ground feature features are solved in a self-training iteration mode. The problems that an existing UNet model is single in structure, complex in parameter and long in interpretation time are solved, and the application range, the interpretation speed and the flexibility degree of the UNet model under different threshold value conditions are increased.

Description

technical field [0001] The invention relates to the technical field of deep learning in remote sensing image target interpretation, in particular to an adaptive perceptron-based remote sensing image intelligent interpretation method and system. Background technique [0002] Remote sensing technology is an important means to study the status and changes of global land cover. In recent years, deep learning technology has developed rapidly, and it can complete tasks such as identification, extraction, and inversion of targets determined by massive data through iterative training. Theoretically, deep learning can extract specific targets through remote sensing imagery and object interpretation data sets, and realize large-scale remote sensing intelligent interpretation. [0003] At present, the most widely used model for intelligent interpretation of remote sensing ground objects using deep learning technology is the fully convolutional neural network (Fully Convolutional Netwo...

AI Technical Summary

Problems solved by technology

[0005] In view of this, the present invention provides a remote sensing image intelligent interpretation method and system based on an adaptive perceptron, which solves the problem that the traditional UNet model does not consider the difference between the front and background of the target feature, and describes the interpretation of the remote sensing feature as Fixed single quest, not matching reality issue

Images