Multicore fusion and spatial Wishart LapSVM-based semi-supervised polarimetric SAR image classification method

Abstract

The invention discloses a multicore fusion and spatial Wishart LapSVM-based semi-supervised polarimetric SAR image classification method, and mainly aims at solving the problem that the classification precision is low as labelled samples of synthetic aperture radar full-polarimetric SAR images are relatively few in existing classification method. The method comprises the following steps of: obtaining a polarization related matrix T; extracting polarization feature vectors of the matrix T and carrying out normalization processing; establishing a training sample set; constructing a Spatial-Wishart manifold regular term; and calculating classification correctness and outputting a polarimetric SAR image classification result. According to the method, the problem that the traditional non-supervision polarimetric SAR image classification is low in correctness is solved, the malpractices of difficulty artificial labelling and high cost caused by requirement of massive label data are avoided; and by jointly utilizing less labelled data and massive un-labelled cheap data, a better classification effect is obtained, so that the method can be used for target classification, detection and recognition of polarimetric SAR images.

Description

technical field [0001] The invention belongs to the technical field of image processing, relates to a polarimetric SAR image classification method, and can be used in technical fields such as ground object classification and target recognition of the polarimetric SAR image. Background technique [0002] Polarimetric SAR (Polarimetric SAR) is a synthetic aperture radar capable of full polarization measurement of the target. It performs full polarization measurement and imaging of the target by measuring and recording the phase difference information of the combined echoes of different polarization states. Polarimetric SAR data contains richer target scattering information, which can more comprehensively express and describe the target, and improve the ability to identify ground objects. At the same time, it has the advantages of all-weather, all-time, and high resolution. It has outstanding advantages in identification, classification and parameter inversion, so it is widely ...

AI Technical Summary

Problems solved by technology

[0005] Since the current supervised and unsupervised classification methods for polarimetric SAR images have certain limitations, it is an urgent task in this technical field to study an effective semi-supervised classification method for polarimetric SAR images.

Images