Polarimetric SAR classification method based on semi-supervised convolutional neural network

A technology of convolutional neural network and classification method, applied in the field of polarimetric SAR classification based on semi-supervised convolutional neural network, can solve the problems of low classification accuracy and large demand of label data

Active Publication Date: 2018-01-09
XIDIAN UNIV
View PDF3 Cites 44 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0006] The technical problem to be solved by the present invention is to provide a polarization SAR classification method based on a semi-supervised convolutional neural network to improve the accuracy of ground object classification and solve the problems of the traditional supervised convolutional neural network. The network has a large demand for labeled samples, and the technical problem of low classification accuracy when having low-label data

Method used

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View more

Image

Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
  • Polarimetric SAR classification method based on semi-supervised convolutional neural network
  • Polarimetric SAR classification method based on semi-supervised convolutional neural network
  • Polarimetric SAR classification method based on semi-supervised convolutional neural network

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0113] Input the polarization SAR ground object simulation data to be classified, see image 3 (a), input the Pauli decomposition map of the polarized SAR and the coherence matrix T of the polarized SAR image, and obtain the label matrix Y according to the ground object distribution information of the polarized SAR image, see image 3 (b), image 3 (b) It is the image directly generated by the label matrix Y. Different color blocks in the image represent different features. The distribution of the same feature in the label matrix is ​​represented by the same category label. The category of features cannot be determined. The label matrix is ​​represented by 0, and the sample is generated from the coherence matrix T of the polarization SAR image and the Pauli decomposition map of the polarization SAR N is the total number of samples, x i Represents the i-th sample.

[0114] Among them, the sample data is randomly extracted into training samples and test samples at a ratio of 1:99 ac...

Embodiment 2

[0128] 1. Experimental conditions

[0129] The hardware platform is: Intel(R)Core(TM)i5-2410M CPU@2.30GHz, RAM 4.00GB;

[0130] The software platform is: MATLAB R2016b;

[0131] The experiment selected 300×270 partial polarized SAR ground objects in Flevoland, the Netherlands, for testing. The number of categories is 6, which are Bare soil, Potato, Beet, Pea, Wheat and Barley. In the experiment, 1% samples of each type are randomly selected as training samples, and the rest are test samples.

[0132] 2. Experimental content and results

[0133] The present invention combines the Softmax classifier to classify the real data of the polarized SAR ground object, and compares it with other deep learning methods under the same experimental setting, where CNN is a convolutional neural network, Figure 4 (c) is by CNN Figure 4 (a) The result map of classification; the deep belief network WDBN based on Wishart RBM is also used in the experiment, Figure 4 (d) Use WDBN method to Figure 4 (...

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to view more

PUM

No PUM Login to view more

Abstract

The invention discloses a polarimetric SAR classification method based on a semi-supervised convolutional neural network. According to the method, first, image data is input, and super pixel segmentation is performed; training samples and test samples are extracted; a neighborhood preserving and sparse filtering deep learning network (NDSFN) is adopted to perform unsupervised pre-training on parameters of the convolutional neural network, and a needed filter set is obtained; a feature map is obtained through convolution operation; fuzzy processing is performed on the feature map through downsampling; parameter settings of the NDSFN are adjusted according to the filter size in a second convolution layer to obtain a new feature map; a Softmax classifier is utilized to perform image classification on the training samples; and a small quantity of marked samples are adopted for slight adjustment to obtain a polarimetric SAR terrain classification. Through the method, by constructing a novelsemi-supervised convolutional neural network (SNCNN) model, the problem that a traditional supervised convolutional neural network has a high requirement for marked samples is effectively solved.

Description

Technical field [0001] The invention belongs to the technical field of image processing, and specifically relates to a polarization SAR classification method based on a semi-supervised convolutional neural network, which can be used for environmental monitoring, earth resource surveying, military systems and the like. Background technique [0002] Machine Learning (ML) is a multi-field interdisciplinary subject, involving probability theory, statistics, approximation theory, convex analysis, algorithm complexity theory and other subjects. Specializing in the study of how computers simulate or realize human learning behaviors in order to acquire new knowledge or skills, and reorganize the existing knowledge structure to continuously improve its own performance. In the field of polarization SAR image classification, machine learning has made many breakthroughs, such as Wishart maximumlikelihood (WML), support vector machines (SVM) and other methods. [0003] Common machine learning ...

Claims

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to view more

Application Information

Patent Timeline
no application Login to view more
IPC IPC(8): G06K9/62G06N3/04G06N3/08
Inventor 刘红英王志杨淑媛焦李成慕彩虹熊涛王桂婷冯婕朱德祥
Owner XIDIAN UNIV
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Try Eureka
PatSnap group products