Looking for breakthrough ideas for innovation challenges? Try Patsnap Eureka!

Semi-supervised polarized SAR image classification method based on random forest composition

A technology of random forest and classification method, applied in the field of image processing, can solve the problems of not considering the spatial information of image sample points, poor classification effect, and inability to accurately represent the structure relationship of SAR data, so as to reduce the use of labeled samples, The effect of similarity relationship graph affinity and improving classification accuracy

Active Publication Date: 2017-11-17
XIDIAN UNIV
View PDF4 Cites 22 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0004] In the construction of the above-mentioned graphical model, the strength of the similarity relationship between sample points in the image is usually expressed by calculating the distance between two data points. However, this simple method of calculating the distance between data points cannot It accurately represents the structural relationship between SAR data with nonlinear structure, and does not consider the spatial information between image sample points, resulting in poor classification effect

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
  • Semi-supervised polarized SAR image classification method based on random forest composition
  • Semi-supervised polarized SAR image classification method based on random forest composition
  • Semi-supervised polarized SAR image classification method based on random forest composition

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0026] Due to the development of remote sensing technology, it has been widely used in environmental monitoring, earth resource survey, military system and other fields, and the demand for polarimetric SAR image processing is also increasing. Some existing polarimetric SAR classification methods include supervised Classification methods, the classification accuracy rate is low, and there are some graph-based semi-supervised methods, mostly by calculating the distance between image sample points to construct a similarity graph matrix, but polarimetric SAR data is nonlinear, this The composition method similarity relationship is not accurate. Therefore, the present invention proposes a semi-supervised classification method for polarization SAR images based on random forest composition, see figure 1 , including the following steps:

[0027](1) Input image: Get the raw data of the polarimetric SAR image from the polarimetric SAR image data folder. In practical applications, use ...

Embodiment 2

[0038] The polarized SAR image semi-supervised classification method based on random forest composition is the same as embodiment 1, and the process of training semi-supervised random forest in step (4) of the present invention is as follows:

[0039] 4a. Initialize the number of training iterations S=0, select s label samples from the label samples to train two classifiers respectively, the value of s depends on the actual size of the image, in this embodiment s=80;

[0040] 4b. Use the first attribute set X 1 The first KNN classifier f is trained on 80 labeled samples selected from 1 , with the second attribute set X 2 The 80 label samples selected in the training second KNN classifier f 2 ;

[0041] 4c. Use these two classifiers to assist the random forest model in semi-supervised training. The use of two classifiers to assist the random forest model in training is to ensure the efficiency of training, which is more efficient and accurate than training with one classifi...

Embodiment 3

[0044] The polarized SAR image semi-supervised classification method based on random forest composition is the same as embodiment 1-2, and the process of optimizing semi-supervised random forest in step (5) is as follows:

[0045] 5a. The present invention uses a deterministic annealing process (Deterministic Annealing process, DA) to perform optimization processing, by introducing a class label distribution probability of unlabeled data in Adjusted for normalization. Add the part of unlabeled data to the optimization goal, the expression is as follows:

[0046]

[0047] The first part of the sum term For the labeled data loss, the second term For the expectation of unlabeled data loss, the third term Expressed as the information entropy of the unlabeled data distribution; α is the expected weight value of the unlabeled data loss and α∈[0,1], the value is 0.5 in this embodiment, T is the temperature variable of annealing, in this embodiment The initial value in is...

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 semi-supervised polarized SAR image classification method based on random forest composition. The method mainly solves a problem that a conventional classification method has a defect in representation of a similarity relation between sample points and does not utilize spatial information. The method comprises the following steps of: inputting raw data of a polarized SAR image; extracting relevant features of the data to obtain a data set; constructing an initial random forest model; training two classifiers by using two different sample sets with different attributes to help to train a semi-supervised random forest mode; optimizing the semi-supervised random forest model; constructing a similarity relation graph; constructing a spatial information graph; combining the similarity relation graph and the spatial information graph to obtain a similarity relation matrix between the sample points; and classifying the images and calculating a correct rate. The method constructs an amiable similarity relation graph and spatial information by using the semi-supervised random forest algorithm, improves the classification accuracy of the polarized SAR image, and can be used in civilian and military fields such as geological exploration, disaster relief, target identification and the like.

Description

technical field [0001] The invention belongs to the technical field of image processing, and mainly relates to classification of polarimetric SAR images, in particular to a semi-supervised classification method of polarimetric SAR images based on random forest composition, which can be used for object classification and target recognition of polarimetric SAR images. Background technique [0002] Polarization SAR is a microwave imaging radar that utilizes the principle of synthetic aperture to achieve high resolution. Texture features and obvious geometric structures of ground objects can be widely used in many fields such as military affairs, agriculture, navigation, and geographical surveillance. It is highly valued in the field of international remote sensing, so polarimetric SAR image classification has become an important research direction of polarimetric SAR information processing. [0003] The purpose of polarimetric SAR image classification is to use the polarizatio...

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
Patent Type & Authority Applications(China)
IPC IPC(8): G06K9/00G06K9/62
CPCG06V20/13G06F18/24323
Inventor 刘红英杨淑媛邢兴慕彩红焦李成缑水平王爽侯彪
Owner XIDIAN UNIV
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Patsnap Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Patsnap Eureka Blog
Learn More
PatSnap group products

Browse by: Latest US Patents, China's latest patents, Technical Efficacy Thesaurus, Application Domain, Technology Topic, Popular Technical Reports.

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap|About US| Contact US: help@patsnap.com