基于多特征融合的层次支持向量机遥感图像云检测

大气与环境光学学报 ›› 2021, Vol. 16 ›› Issue (1): 58-66.

基于多特征融合的层次支持向量机遥感图像云检测

张波1;2, 胡亚东1∗, 洪津1

1 中国科学院合肥物质科学研究院安徽光学精密机械研究所, 中国科学院通用光学定标与表征技术重点实验室, 安徽合肥 230031; 2 中国科学技术大学, 安徽合肥 230026

收稿日期:2020-02-23 修回日期:2020-04-28 出版日期:2021-01-28 发布日期:2021-02-02
通讯作者: E-mail: huyadong@aiofm.ac.cn E-mail:huyadong@aiofm.ac.cn
作者简介:张波 (1990- ), 安徽马鞍山人, 硕士, 主要从事遥感图像处理方面的研究。 E-mail: zhangbo1@mail.ustc.edu.com
基金资助:
Supported by K.C.Wong Education Foundation “International Team of Advanced Polarization Remote Sensing Technology and Application” (王宽诚教育基金“先进偏振遥感技术与应用国际团队”, GJTD-2018-15)

Cloud Detection of Remote Sensing Images Based on H-SVM with Multi-Feature Fusion

¹, ¹

1 Key Laboratory of Optical Calibration and Characterization, Anhui Institute of Optics and Fine Mechanics, HFIPS, Chinese Academy of Sciences, Hefei 230031, China; 2 University of Science and Technology of China, Hefei 230026, China

Received:2020-02-23 Revised:2020-04-28 Published:2021-01-28 Online:2021-02-02

摘要/Abstract

摘要： 云检测是遥感图像处理和应用的前提, 针对遥感图像云检测的准确率容易受到薄云及似云地物影响的挑战, 提出一种结合遥感影像灰度、纹理和频率特征的层次支持向量机云检测算法。该方法首先采用简单线性迭代聚类算法将遥感图像分割为像素块, 再采用一种层次支持向量机分类器对遥感图像以像素块为单位进行云检测。层次支持向量机的第一层将像素块初步分为 “云” 和 “地物” 两类。层次向量机的第二层针对第一层分类的结果分别设计两个分类器进行进一步分类, 并将分类后的结果合并为 “厚云”、 “薄云”、 “地物” 三类。最后, 将分类结果进行膨胀处理, 得到最终的云检测结果。选取高分一号 WFV 的 RGB 波段遥感图像进行实验, 结果显示提出的新方法对实验图像的云检测平均准确率为 95.4%, 表明该方法可适用于多种场景下遥感图像的云检测, 服务于遥感产品的生产和应用。

关键词: 云检测, 层次支持向量机, 简单线性迭代聚类, 多特征融合

Abstract: Cloud detection is the prerequisite of remote sensing image processing and application. It is a widely challenge that the accuracy of cloud detection from remote sensing image is easily influenced by thin clouds and cloud-like ground targets. Therefore, a hierarchical support vector machine (H-SVM) cloud detection algorithm combining grayscale, texture, and frequency features of remote sensing image is proposed in this work. Firstly, a simple linear iterative clustering algorithm is used to segment the remote sensing image into pixel blocks. Secondly, a H-SVM classifier is designed to perform cloud detection on the segmented pixel blocks, where the first layer of the H-SVM preliminarily divides the pixel blocks into “cloud” and “landscape categories”, and the second layer containing two classifiers further classifies the classification results of the first layer and then merges the classified results into three categories of “thick cloud”, “thin cloud”, and “land features”. Finally, the classification results are processed using expansion algorithm to get the final cloud detection results. RGB band remote sensing images of GF-1 WFV are selected for verification experiments. It is shown that the method proposed in this study has an average accuracy of 95.4% for the cloud detection in the experimental images, which indicates that the method can be used for cloud detection of remote sensing images in multiple scenarios, and serve the production and application of remote sensing products.

Key words: cloud detection, hierarchical support vector machine, simple linear iterative clustering, multi-feature fusion

中图分类号:

TP75

张波, 胡亚东∗, 洪津. 基于多特征融合的层次支持向量机遥感图像云检测[J]. 大气与环境光学学报, 2021, 16(1): 58-66.

参考文献 99

[1]	Hou Shuwei, Sun Wenfang, Zheng Xiaosong. Overview of cloud detection methods in remote sensing images [J]. Space
	Electronic Technology, 2014, 11(3): 68-76, 86.
	侯舒维, 孙文方, 郑小松. 遥感图像云检测方法综述 [J]. 空间电子技术, 2014, 11(3): 68-76, 86.
[2]	Wang Quan, Sun Lin, Wei Jing, et al. Improvement of universal dynamic threshold cloud detection algorithm and its application
	in high resolution satellite [J]. Acta Optica Sinica, 2018, 38(10): 376-385.
	王权, 孙林, 韦晶, 等. 动态阈值云检测算法改进及在高分辨率卫星上的应用 [J]. 光学学报, 2018, 38(10): 376-385.
[3]	Jia Liangliang, Wang Xiaoqin, Wang Feng. Cloud detection based on band operation texture feature for GF-1 multispectral
	data [J]. Remote Sensing Information, 2018, 33(5): 62-68.
	贾亮亮, 汪小钦, 王峰. 基于波段运算和纹理特征的高分一号多光谱数据云检测 [J]. 遥感信息, 2018, 33(5): 62-68.
[4]	Gao Jun, Wang Kai, Tian Xiaoyu, et al. A BP-NN based cloud detection method for FY-4 remote sensing images [J]. Journal
	of Infrared and Millimeter Wave, 2018, 37(4): 477-485.
	高军, 王恺, 田晓宇, 等. 基于 BP 神经网络的风云四号遥感图像云检测算法 [J]. 红外与毫米波学报, 2018, 37(4):
47	7-485.
[5]	Li P F, Dong L M, Xiao H C, et al. A cloud image detection method based on SVM vector machine [J]. Neurocomputing,
20	15, 169: 34-42.
[6]	Kang Yifei. Methods and Applications of Cloud Detection for Optical Remote Sensing Satellite Imagery [D]. Wuhan: Wuhan
	University, 2018.
	康一飞. 光学遥感卫星影像云检测方法及应用 [D]. 武汉: 武汉大学, 2018.
[7]	Xu Jiayao． Haze Removal Research for GF-1 Satellite Image [D]. Beijing: China University of Geosciences (Beijing), 2015.
	许佳垚. 高分一号卫星影像去薄云方法研究 [D]. 北京: 中国地质大学(北京), 2015．
[8]	Achanta R, Shaji A, Smith K, et al. SLIC superpixels compared to state of the art superpixel methods [J]. IEEE Transactions
	on Pattern Anal Machine Intelligence, 2012, 34(11): 2274-2281.
[9]	Kim K S, Zhang D, Kang M C, et al. Improved simple linear iterative clustering super pixels [C]. Consumer Electronics
	(ISCE), 2013 IEEE 17th International Symposium on. IEEE, 2013.
[10]	Zheng Hong, Li Zhao, Li Jun. Study on fast implementation and optimal method of gray level co-occurrence matrix [J].
	Chinese Journal of Scientific Instrument, 2012, 33(11): 2509-2515.
	郑红, 李钊, 李俊. 灰度共生矩阵的快速实现和优化方法研究 [J]. 仪器仪表学报, 2012, 33(11): 2509-2515.
[11]	Nanni L, Lumini A, Brahnam S. Survey on LBP based texture descriptors for image classification [J]. Expert Systems with
	Applications, 2012, 39(3): 3634-3641.
[12]	Dan Z P, Sang N, He Y G, et al. An improved LBP transfer learning for remote sensing object recognition [J]. OptikInternational Journal for Light and Electron Optics, 2014, 125(1): 482-485.
[13]	Wang Shiyao. Research on Thin Cloud Removal for Single Aerial Images [D]. Wuhan: Wuhan University, 2017.
	王诗尧. 单幅遥感图像去薄云算法研究 [D]. 武汉: 武汉大学, 2017.
[14]	Chen Fen, Yan Dongmei, Zhao Zhongming. Haze detection and removal in remote sensing images based on undecimated
	wavelet transform [J]. Geomatics and Information Science of Wuhan University, 2007, 32(1): 71-74.
	陈奋, 闫冬梅, 赵忠明. 基于无抽样小波的遥感影像薄云检测与去除 [J]. 武汉大学学报 (信息科学版), 2007, 32(1):
71	-74.
[15]	Zhang R, Sun D L, Li S M, et al. A stepwise cloud shadow detection approach combining geometry determination and SVM
	classification for MODIS data [J]. International Journal for Remote Sensing, 2013, 34(1): 211-226.
[16]	Du Zhaohui, Si Yujuan. Emotion recognition of pulse signal based on hierarchical support vector machine [J]. Journal of Jilin
	University (Information Science Ed), 2017, 35(1): 37-42.
	杜昭慧, 司玉娟. 基于层次支持向量机的脉搏信号情感识别 [J]. 吉林大学学报 (信息科学版), 2017, 35(1): 37-42.
[17]	Dong Zhipeng, Wang Mi, Li Deren, et al. A cloud detection method for high resolution remote sensing imagery based on the
	spectrum and texture of objects [J]. Acta Geodaetica et Cartographica Sinica, 2018, 47(7): 996-1006.
	董志鹏, 王密, 李德仁, 等. 利用对象光谱与纹理实现高分辨率遥感影像云检测方法 [J]. 测绘学报, 2018, 47(7):
99	6-1006.