SAR Image Classification Method Based on Depth Correlation Vector Machine

Abstract

The invention discloses a SAR image classification method based on a depth correlation vector machine, which mainly solves the problem of synthetic aperture radar image classification. The classification process is: (1) input the matrix of double-precision SAR SAR images; (2) normalize the input matrix of double-precision SAR SAR images; (3) train three Restricted Boltzmann machine RBM; (4) training RVM classifier; (5) constructing deep RVM model; (6) classification; (7) calculating prediction class label; (8) obtaining classification accuracy. The present invention excavates the depth feature of the synthetic aperture radar image, adopts the correlation vector machine RVM, learns the feature sparsely, retains the information integrity of the radar image, mines the depth information, reduces the time complexity, and has a good classification effect, It can be used for SAR image classification.

Description

Technical field [0001] The present invention belongs to the technical field of image processing, and further relates to a synthetic aperture radar (Synthetic Aperture Radar, SAR) image classification method based on a depth correlation vector machine (Relevance Vector Machine, RVM) in the technical field of image classification. The invention can be applied to target classification and recognition of SAR images. Background technique [0002] In recent years, with the rapid improvement of computer performance and the rapid development of the Internet, how to make better use of statistical methods, machine learning, data mining and other technologies to extract valuable information from them to deal with practical problems in production and life has increasingly become an issue. Important and urgent research topics. Among them, classification is the most basic and important data processing technology. It has a wide range of fields such as automatic text classification, mass web pa...

AI Technical Summary

Problems solved by technology

However, the shortcomings of this method are that it only has a good classification effect on some data sets and cannot be transplanted to other types of data sets, and its robustness is not strong, which reduces the performance of the algorithm.

This method uses the data information of the image, the correlation information between the image and the relevant information of the image on multiple scales, and extracts important information from a small amount of weak information, that is, learns SAR from the training data marked with keywords. The image classification model can greatly reduce the difficulty of obtaining accurate training data, and overcome some local uncertainties in SAR image classification. In the process of multiple related information, the information between the data is cut too fast, and a small amount of weak information used for training loses a lot of important details, resulting in low multiple average classification accuracy and insufficient robustness of the classification method.

Images