A polarimetric SAR image classification method, system, device and medium based on agent-assisted evolutionary neural network architecture search

By adopting a proxy-assisted evolutionary neural network architecture search method in polarized SAR image classification, the high cost and low efficiency of artificially designed neural network models are solved, and efficient and high-precision neural network model search and classification effects are achieved.

CN116977854BActive Publication Date: 2025-04-29XIDIAN UNIV
3 Cites 0 Cited by

Patent Information

Application Number
CN202310949167.1
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2023-07-31
Publication Date
2025-04-29
Estimated Expiration
2043-07-31

AI Technical Summary

Technical Problem

The prior art has high cost and artificial error problems in the polarized SAR image classification, and the search method for evolutionary neural network architecture is time-consuming and resource-consuming, resulting in low model search efficiency.

Method used

Using a proxy-assisted evolutionary neural network architecture search method, the architectural parameters of complex-valued neural network are encoded, and the genetic algorithm is used to combine the proxy model to select the optimal complex-valued convolutional neural network to achieve efficient and high-precision classification of polarized SAR images.

Benefits of technology

It reduces the time cost and resource consumption of the evolutionary architecture search algorithm, and realizes efficient and high-precision search of neural network models suitable for polarized SAR image classification, reducing the need for manual design and artificial error.

✦ Generated by Eureka AI based on patent content.
Patent Text Reader

Abstract

A polarimetric SAR image classification method, system, device and medium based on agent-assisted evolutionary neural network architecture search, the method comprising: complex-valued neural network model encoding, agent model training database data sampling, agent model training database data preprocessing, agent model training, obtaining an optimal complex-valued convolutional neural network, and SAR image classification; the system, device and medium are used to implement a polarimetric SAR image classification method based on agent-assisted evolutionary neural network architecture search; the present invention can reduce the time cost and resource consumption of the evolutionary architecture search algorithm, and can efficiently and accurately search for a neural network model suitable for polarimetric SAR image classification.
Need to check novelty before this filing date? Find Prior Art