|
|
- 2016
基于人类感知的SAR图像海上溢油检测算法
|
Abstract:
在基于SAR图像的海上溢油检测中,识别的效率与准确率是关键。纹理特征是人类专家能够较好的判别SAR图像中油膜,类油膜以及海水的一个重要依据。本文算法一方面融合灰度共生矩阵与Tamura特征,直接对SAR原始图像进行特征提取,避免了对图像进行分割、降噪等预处理,提高了识别算法的可行性与识别效率。另一方面,应用深度信念网络(DBN)的分类方法,可以很好地解决溢油检测中小样本分类的问题,并且模仿人类感知系统高效准确的表示信息、获取本质特征。本文应用人类感知的思想对油膜、类油膜以及海水这3类样本进行分类识别。通过实验确定了DBN中利于分类的关键参数值。本算法对原始SAR图像中3类样本的识别准确率达到90.36%,具有较好的实用价值
| [1] | Tamura H,Mori S,Yamawaki T. Textural Features Corresponding to Visual Perception[J].<em> IEEE Transaction on Systems,</em>1978,8(6):460-473 |
| [2] | Bengio Y,Lecun Y. Scaling Learning Algorithms Towards AI[M]. Cambridge: MIT Press,2007: 321-358; 599-619 |
| [3] | Wang Jun. Estimation of the Kappa Coefficient from Consistency[D]. Chengdou: Sichuan University, 2006(王军. Kappa系数在一致性评价中的应用研究[D].成都:四川大学,2006) |
| [4] | Wu Zhaocong,Ouyang Qundong,Hu Zhongwen. Polarimetric SAR Image Classification Using Watershed-Transformation and Support Vector Machine [J]. <em>Geomatics and Information Science of Wuhan University</em>,2012,37(1):7-10(巫兆聪,欧阳群东,胡忠文.应用分水岭变换与支持向量机的极化SAR图像分类[J].武汉大学学报·信息科学版, 2012,37(1):7-10) |
| [5] | Zhou Shusen, Chen Qingcai, Wang Xiaolong. Active Deep Learning Method for Semi-supervised Sentiment Classification[J]. <em>Neurocomputing,</em> 2013, 120: 536-546 |
| [6] | Ma Long,Li Ying,Niu Ying. Research on SAR Oil Spill Monitoring by SVM Based on Texture Features[J]. <em>Navigation of China</em>, 2010,33(1):75-79(马龙,李颖,牛莹. 结合纹理的支持向量机合成孔径雷达溢油监测[J]. 中国航海,2010,33(1):75-79) |
| [7] | Shi Lijian,Zhao Chaofang,Liu Peng. Oil Spill Identification in Marine SAR Images Based on TextureFeature and Artificial Neural Network[J]. <em>Periodical of Ocean University of China</em>,2009,39(6): 1 269-1 274(石立坚,赵朝方,刘朋.基于纹理分析和人工神经网络的SAR图像中海面溢油识别方法[J]. 中国海洋大学学报,2009,39(6):1 269-1 274) |
| [8] | Mansor S B, Assilzadeh H, IbrahimH M, et al. Oil Spill Detection and Monitoring from Satellite Image [OL]. http:// www.Gisdevelopment.net/appli2cation/miscellaneous,2014 |
| [9] | Utgoff P,Stracuzzi D. Many Layered Learning[J].<em>Neural Computation</em>,2002,14(10):2 497-2 529 |
| [10] | Topouzelis K,Karathanassi V,Pavlakis P,et al. Oil Spill Detection Using Netural Netwrks and SAR Data[C].The 20th ISPRS Congress,Istanbul,2004 |
| [11] | Zhao Dakang. Synthetic Aperture Radar Image Classification Based on Neural Networks[J]. <em>Science Information</em>, 2006(7):5-8(赵大康.基于神经网络的合成孔径雷图像分类研究[J].科技信息,2006(7):5-8) |
| [12] | He Chu,Liu Ming. A Hierarchical Classification Method Based on Feature Selection and Adaptive Decision Tree for SAR Image[J]. <em>Geomatics and Information Science of Wuhan University</em>,2012,37(1):46-49(何楚,刘明.利用特征选择自适应决策树的层次SAR图像分类[J].武汉大学学报·信息科学版,2012,37(1):46-49) |
| [13] | Bengio Y,Lecun Y. Scaling Learning Algorithms Towards AI[J].<em>Large-Scale Kernel Machines</em>,2007,1:1-41 |
| [14] | G E. Hinton,S Osindero,Yee W. Teh. A Fast Learning Algorithm for Deep Belief Nets[J]. <em>Neural Comput</em>, 2006,18(7):1 527-1 554 |
| [15] | Bengio Y,Delalleau O. On the Expressive Power of Deep Architectures [C]. The 14th International Conference on Discovery Science, Berlin,2011 |
| [16] | del Frate F, Petrocchi A, Lichtenegger J, et al. Neural Networks for Oil Spill Detection Using ERS-SAR Data[J]. <em>IEEE Transactions on</em> Geoscience and Remote Sensing, 2000,38(5): 2 282-2 287 |
| [17] | Xiao Hanguang, Cai Congzhong. Comparison Study of Normalization of Feature Vector[J]. <em>Computer Engineering and Applications</em>,2009,45(22):117-119(肖汉光,蔡从中. 特征向量的归一化比较性研究[J]. 计算机工程与应用,2009,45(22):117-119) |
| [18] | Lena Chang, Tang Z S, Chang S H, et al. A Region-based GLRT Detection of Oil Spills in SAR Images[J].<em> Pattern Recognition Letters,</em> 2008,29: 1 915-1 923 |
| [19] | Konstantinos T, Apostolos P. Oil Spill Feature Selection and Classification Using Decision Tree Forest on SAR Image Data[J]. <em>ISPRS Journal of Photogrammetry and Remote Sensing,</em>2012, 68: 135-143 |
