Polarization SAR Image Classification Method Based on Scattered Energy and Stack Autoencoder

Abstract

The invention discloses a scattering energy and stack self-code-based polarimetric SAR image classification method, and mainly aims at solving the problems that classification process is complicated,polarimetric SAR image spatial features cannot be extracted and classification precision is not high due to the influences of extraction data independence and redundancy of polarimetric synthetic aperture radar SAR image features in the prior art. The method comprises the following specific realization steps of: (1) inputting a coherence matrix of a to-be-classified polarimetric synthetic apertureradar SAR image; (2) compensating the polarimetric SAR coherence matrix; (3) obtaining scattering energy of a scanning model; (4) obtaining a sample; (5) training stack self-codes; (6) classifying the stack self-codes; and (7) outputting a classification result. The method has the advantage of being remarkable in polarimetric synthetic aperture radar SAR image classification effect.

Description

technical field [0001] The invention belongs to the technical field of image processing, and further relates to a polarization synthetic aperture radar (Synthetic Aperture Radar SAR) image classification method based on scattered energy and stack self-encoding in the field of target recognition. The invention can be used for ground object classification and target recognition of polarization synthetic aperture radar SAR images. Background technique [0002] Compared with traditional synthetic aperture radar, polarimetric synthetic aperture radar (SAR) uses the scattering information of multiple channels to obtain a more comprehensive understanding of the target. The classification of polarimetric SAR SAR images is an important research content of polarimetric SAR SAR image interpretation. The classification map can be used as an intermediate result to provide auxiliary information for edge extraction, target detection, recognition, etc., and can also be directly output as th...

AI Technical Summary

Problems solved by technology

Although this method combines the texture characteristics of the power image, it still has the disadvantage that when using the deep neural network to extract the features of the polarimetric synthetic aperture radar SAR image, it fails to introduce the spatial neighborhood features, which leads to unreasonable feature extraction and greatly affects Classification Accuracy of Synthetic Aperture Radar SAR Images

Images