A SAR image classification method based on an improved OGMRF-RC model
LI Yuan1,2,3,4(), WU Lin1,2,3,5, QI Wenwen1,2,3, GUO Zhengwei1,2,3(), LI Ning1,2,3
1. College of Computer and Information Engineering, Henan University, Kaifeng 475004, China 2. Henan Engineering Research Center of Intelligent Technology and Application, Henan University, Kaifeng 475004, China 3. Henan Key Laboratory of Big Data Analysis and Processing, Henan University, Kaifeng 475004, China 4. School of Information and Electronic Engineering, Shangqiu Institute of Technology, Shangqiu 476000, China 5. College of Environment and Planning, Henan University, Kaifeng 475004, China
The classification of synthetic aperture Radar (SAR) images is one of the key technologies in the field of remote sensing applications. To address the problem that regional class labels affect the classification accuracy in the object-based Markov random field (OMRF) model, this paper proposes the concept of regional category fuzzy probability (RCFP) label field, which can effectively avoid the misclassification caused by wrong class labels by fully considering the possible categories of a single object. The RCFP of every region can be obtained using the regional edge information and posterior probability according to the features of the region and its adjacent regions. Then it is included in the calculation of feature field parameters to make the feature field parameters highly close to the real conditions of objects. The study area is located in the eastern part of Kaifeng City, Henan Province, covering an area of about 1 400 km2. Sentinel-1 SAR images were used for the classification experiment of farmlands, buildings, and water in the study area, and the performance of the improved method in this study was compared with that of the method of K-means, fuzzy C-means (FCM), MRF, and OGMRF-RC. The experimental results show that the overall accuracy (OA) and the Kappa coefficient of the proposed method are 94.16% and 0.8957 respectively, which are higher than those of other methods.
李渊, 毋琳, 戚雯雯, 郭拯危, 李宁. 基于改进OGMRF-RC模型的SAR图像分类方法[J]. 自然资源遥感, 2021, 33(4): 98-104.
LI Yuan, WU Lin, QI Wenwen, GUO Zhengwei, LI Ning. A SAR image classification method based on an improved OGMRF-RC model. Remote Sensing for Natural Resources, 2021, 33(4): 98-104.
Fan W B, Sun Z Y. SAR image segmentation algorithm based on generalized gauss distribution in wavelet domain[J]. Computer Engineering and Applications, 2020,56(5):222-226.
Lai Q, Zhang J Q, Xi Y Y, et al. A ship detection method for SAR imagery based on three-centroid K-Means clustering algorithm[J]. Journal of Northeast Normal University(Nature Science Edition), 2017,49(3):61-67.
Guo X Y, Li K, Wang Z Y, et al. Fine classification of rice with multi-temporal compact polarimetric SAR based on SVM+SFS strategy[J]. Remote Sensing for Land and Resources, 2018,30(4):20-27.doi: 10.6046/gtzyyg.2018.04.04.
doi: 10.6046/gtzyyg.2018.04.04
[5]
Voisin A, Krylov V A, Moser G, et al. Classification of very-high resolution SAR images of urban areas using copulas and texture in a hierarchical markov random field model[J]. IEEE Geoscience and Remote Sensing Letters, 2013,10(1):96-100.
doi: 10.1109/LGRS.2012.2193869
[6]
Delia C, Ruscino S, Abbate M, et al. SAR image classification through information-theoretic textural features,MRF segmentation,and object-oriented learning vector quantization[J]. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, 2014,7(4):1116-1126.
doi: 10.1109/JSTARS.4609443
[7]
Shen Y, Chen J Y, Xiao L, et al. A bilevel contextual MRF model for supervised classification of high spatial resolution remote sensing images[J]. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, 2019,12(12):360-5372.
[8]
Bi H X, Xu L, Cao X, et al. Polarimetric SAR image semantic segmentation with 3D discrete wavelet transform and Markov random field[J]. IEEE Transactions on Image Processing, 2020,29:6601-6614.
doi: 10.1109/TIP.2020.2992177
[9]
Jiang X F, Zhang Y, Li Y, et al. Hyperspectral image classification with transfer learning and Markov random fields[J]. IEEE Geoscience and Remote Sensing Letters, 2020,17(3):544-548.
doi: 10.1109/LGRS.8859
[10]
Lee S H. Hierarchical classification of land-cover types using RAG-based merging[C]//2007 IEEE International Geoscience and Remote Sensing Symposium, 2007:2060-2062.
[11]
Chen X H, Zheng C, Yao H T, et al. Image segmentation using a unified Markov random field model[J]. The Institution of Engineering and Technology, 2017,11(10):860-869.
[12]
Zheng C, Yao H T. Segmentation for remote-sensing imagery using the object-based Gaussian-Markov random field model with region coefficients[J]. International Journal of Remote Sensing, 2019,40(11):4441-4472.
doi: 10.1080/01431161.2018.1563841
[13]
Xu X F, Zou B, Zhang L M. PolSAR image classification based on object-based Markov random field with polarimetric auxiliary label field[J]. IEEE Geoscience and Remote Sensing Letters, 2020,17(9):1558-1562.
doi: 10.1109/LGRS.8859
Li Z, Qu C W, Zhou Q, et al. A sea-land segmentation algorithm of SAR image based on the SLIC superpixel division[J]. Radar Science and Technology, 2017,15(4):354-358.