面向对象的 PolSAR 图像 SVM 分类

中国民航大学学报 ›› 2022, Vol. 40 ›› Issue (1): 21-26.

面向对象的 PolSAR 图像 SVM 分类

韩宾宾，韩萍，程争

（中国民航大学 a. 电子信息与自动化学院；b. 工程技术训练中心，天津 300300）

收稿日期:2021-10-20 修回日期:2021-10-20 接受日期:2021-09-06 出版日期:2022-02-22 发布日期:2022-03-17
作者简介:韩宾宾（1991—），男，河北保定人，助理实验师，硕士，研究方向为 PolSAR 图像解译、计算机视觉。
基金资助:
中央高校基本科研业务费专项（3122019046）

Object-oriented SVM classification of PolSAR image

HAN Binbin, HAN Ping , CHENG Zheng

a. College of Electronic Information and Automation; b. Engineering Techniques Training Center， CAUC， Tianjin 300300, China

Received:2021-10-20 Revised:2021-10-20 Accepted:2021-09-06 Online:2022-02-22 Published:2022-03-17

摘要/Abstract

摘要： 极化合成孔径雷达（PolSAR，polarimetric synthetic aperture radar）图像具有强相干斑噪声和大场景特点，为此提出一种面向对象的支持向量机（SVM， support vector machine）分类算法。算法首先通过超像素分割产生待分类对象，以此减少分类处理单元，同时实现特征滤波降噪；然后通过转换矩阵提取信息完备且具有简单统计描述的雷达散射截面积特征；最后，选择在小样本条件下仍具有较强学习能力和泛化能力的 SVM分类器实现图像分类。用公开的实测 San Francisco 数据进行实验，实验结果表明：该算法相对于对比算法在准确率上提升约 10%。

关键词: 极化合成孔径雷达（PolSAR）, 面向对象, 支持向量机（SVM）, 超像素分割

Abstract: Polarimetric synthetic aperture radar(PolSAR) image is characterized with strong speckle noise and big scene. Therefore, an object-oriented support vector machine(SVM) classification algorithm was proposed. Firstly, the object to be classified is generated by superpixel segmentation, which reduces the classification unit and achieves noise reduction; Then, the radar cross section (RCS) features with complete information and simple statistical description are extracted by the conversion matrix; Finally, SVM classifier with strong learning ability and generalization ability is selected to realize image classification. Experiments are conducted with publicly measured San Francisco data, and the experimental results show that the accuracy of this algorithm is improved by 10% compared with the contrast algorithm.

Key words: polarimetric synthetic aperture radar(PolSAR), object-oriented, support vector machine(SVM), superpixel segmentation

中图分类号:

韩宾宾 , 韩萍 , 程争 . 面向对象的 PolSAR 图像 SVM 分类[J]. 中国民航大学学报, 2022, 40(1): 21-26.

HAN Binbin, HAN Ping , CHENG Zheng. Object-oriented SVM classification of PolSAR image [J]. Journal of Civil Aviation University of China, 2022, 40(1): 21-26.

参考文献 10

[1]	聂祥丽, 黄夏渊, 张波, 等．极化合成孔径雷达图像相干斑抑制和分类方法综述[J]．自动化学报, 2019, 45(8): 1419-1438．
[2]	ZHANG L M, ZOU B, ZHANG J P, et al． Classification of polarimetric SAR image based on support vector machine using multiple-component scattering model and texture features[J]． EURASIP Journal on Advances in Signal Processing, 2010: 960831．
[3]	张祥, 邓喀中, 范洪冬, 等．基于目标分解的极化 SAR 图像 SVM 监督分类[J]．计算机应用研究, 2013, 30(1): 295-298．
[4]	XING Y X, ZHANG Y, LI N, et al． Improved superpixel-based polarimetric synthetic aperture radar image classification integrating color features[J]. Journal of Applied Remote Sensing, 2016, 10(2): 026026.
[5]	WANG H J, HAN J H, DENG Y Y． PolSAR image classification based on Laplacian Eigenmaps and superpixels[J]． EURASIP Journal on Wireless Communications and Networking, 2017(1): 1-9．
[6]	韩景红, 王海江．基于 MV 与 Wishart 距离的极化 SAR 图像分类[J]. 成都信息工程大学学报, 2019, 34(1): 31-34．
[7]	HAN B B, HAN P, CHENG Z． A fast and accurate cluster center initialization algorithm for PolSAR superpixel segmentation[C]//2019 Asia-Pacific Signal and Information Processing Association Annual Summit and Conference(APSIPAASC), November 18-21, 2019, Lanzhou, China． IEEE, 2019: 2032-2037．
[8]	HOEKMAN D H, VISSERS M A M． A new polarimetric classification approach evaluated for agricultural crops[J]． IEEE Transactions on Geoscience and Remote Sensing, 2003, 41(12): 2881-2889．
[9]	LEE Y J, MANGASARIAN O L． SSVM: A smooth support vector machine for classification[J]． Computational Optimization and Applications, 2001, 20(1): 5-22．
[10]	XIAO D L, LIU C． PolSAR image classification based on dilated convolution and pixel-refining parallel mapping network in the complex domain[EB/OL]． (2019-09-24)[2021-09-06]. https://arxiv.org/abs/1909. 10783v2.