Automatic Change Detection of High Resolution Remote  Sensing Images Based on Level Set Evolution and Support Vector Machine Classification

YAN Ming; CAO Guo; XIA Meng

doi:10.15938/j.jhust.2019.01.013

Chinese

您当前的位置：

首页 >

文章列表页 >

Automatic Change Detection of High Resolution Remote Sensing Images Based on Level Set Evolution and Support Vector Machine Classification

更新时间：2026-08-31

- Automatic Change Detection of High Resolution Remote Sensing Images Based on Level Set Evolution and Support Vector Machine Classification
- Vol. 24, Issue 1, Pages: 78-84(2019)
- 作者机构：
  
  南京理工大学计算机科学与技术学院,江苏,南京,210049
- 作者简介：
- 基金信息：
- DOI：10.15938/j.jhust.2019.01.013
  CLC： TP752
- Published：2019
- 稿件说明：
移动端阅览
YAN Ming, CAO Guo, XIA Meng. Automatic Change Detection of High Resolution Remote Sensing Images Based on Level Set Evolution and Support Vector Machine Classification[J]. 2019, 24(1): 78-84.
DOI：

YAN Ming, CAO Guo, XIA Meng. Automatic Change Detection of High Resolution Remote Sensing Images Based on Level Set Evolution and Support Vector Machine Classification[J]. 2019, 24(1): 78-84. DOI： 10.15938/j.jhust.2019.01.013.

摘要

摘要:提出了基于水平集演化和支持向量机（SVM）分类的高分辨率遥感图像变化检测方法，该方法将像素级的和对象级的变化检测方法相结合，运用了像素特征和对象特征以提高变化类和非变化类的准确率。在像素级上，变化检测问题转化为水平集演化的图像分割问题。在对象级上，本文可以从分割结果中为SVM分类器自动地选择潜在的训练样本。最终将基于像素级的变化和基于对象级的变化相结合得到最终的变化结果。所提出的方法的主要优势在于可以自动选择合适的样本进行SVM分类器训练。此外，提出的方法可以有效的提高精确度和自动化水平。通过SPOT5图像和航空图像进行实验，结果表明该方法是有效的。

Abstract

Abstract:We propose a method for change detection in highresolutionremote sensing images by means of level set evolution and Support VectorMachine (SVM) classification

which combined both pixellevel method and objectlevelmethod Both pixelbased change features and objectbased ones are extracted to improvethe discriminability between the changed class and the unchanged classAt thepixellevel

the change detection problem is formulated as a segmentation issueusing level set evolution in the difference image At the objectlevel

potentialtraining samples are selectedfrom the segmentation results without manualintervention into SVM classifier Thereafter

the final changes are obtained bycombining the pixelbased changes and the objectbased changes A chief advantageof our approach is being able to select appropriate samples for SVM classifiertraining Furthermore

our proposed method helps improving the accuracy and thedegree of automation We systematically evaluated it with a variety of SPOT5images and aerial images Experimental results demonstrated the accuracy of ourproposed method

关键词

Keywords

references

Views

403

下载量

CSCD

Alert me when the article has been cited

提交

Tools

Publicity Resources

Image Segmentation Algorithm Named FSICM Based on Spatial Information

Segmentation Algorithm of Spine CT ImageBased on Hidden Markov Ｒandom Field

Nodularization Effect Evaluationof Ductile Iron basedon Quick Testing Oxygen Content in Liquid Iron

Fabrication and Mechanical Property of SiC/SiC Composites by Selective Laser Sintering Technology Combined with Precursor Impregnation and Pyrolysis Process

Stability Analysis of a Class of SIR Epidemic Model with Delayed Spontaneous Variation of Virus

Related Author

ZHU Suxia

ZU Hongliang

SUN Guanglu

LIU Xia

KUANG Ｒui-xue

WANG Bo

YUE Yu-ting

PEI Li-xin

Related Institution

School of Computer Science and Technology, Harbin University of Science and Technology

School of Automation，Harbin University of Science and Technology

School of Material Science and Engineering，Harbin University of Science and Technology

School of Civil Engineering and Architecture, Harbin University of Science and Technology

School of Material Science and Engineering, Harbin University of Science and Technology

AI问答

Postal code：100079
Tel：（010）53879206 Email：tmw@bjxintong.com.cn
Technical support is provided by Beijing Founder electronics co., LTD 京ICP备09082226号-64 京公网安备11010602201714号
It is recommended to read the content of this site in Chrome&IE9+. Please switch to extreme mode in browser 360.
Cookies We use cookies to help provide and enhance our service and tailor content. By continuing, you agree to the use of cookies.

⁰