A semantic segmentation method for Satellite Image Change Detection

被引：0

作者：

Zhang, Jiahao ^{[1
]}

Chen, Bo ^{[1
]}

Zhou, Jianbang ^{[1
]}

Yang, Jingkun ^{[1
]}

Chen, Zhong ^{[1
]}

Yang, Jian ^{[2
]}

Zhang, Yanna ^{[3
]}

机构：

[1] Huazhong Univ Sci & Technol, Natl Key Lab Sci & Technol Multispectral Informat, Inst Artificial Intelligence & Automat, Wuhan, Peoples R China

[2] Chinese Acad Sci, Inst Aerosp Informat Innovat, Beijing, Peoples R China

[3] Henan Univ, Dept Lab & Equipment Management, Kaifeng, Peoples R China

来源：

MIPPR 2019: AUTOMATIC TARGET RECOGNITION AND NAVIGATION | 2020年 / 11429卷

关键词：

change detection; semantic segmentation; convolutional neural network; Deep learning;

D O I：

10.1117/12.2538085

中图分类号：

O43 [光学];

学科分类号：

070207 ; 0803 ;

摘要：

We apply the semantic segmentation method in deep network to high precision satellite image change detection, and propose a network framework to improve the detection performance.We directly processed the image after registration, without the steps of radiometric correction, and avoided the tedious steps of manual feature design by traditional methods.We tried to use Unet and Deeplab v3 model to divide the change area, and added the structure of jumping connection on the basis of Deeplab network, which made the edge of the detection graph more accurate and improved the performance of the network.The test results show that this method is effective for detecting the change of high-precision remote sensing images.

引用

页数：7

共 21 条

[1]

[Anonymous], ARXIV14091556V6

[2]

[Anonymous], 2015, PROC INT C LEARN REP

[3] A minimum-cost thresholding technique for unsupervised change detection [J].