CAJ | 학술논문

为了降低谱聚类算法尺度参数对分类结果的影响，避免 Nystr?m逼近导致的分类结果不稳定，提出了一种基于谱聚类集成的极化合成孔径雷达(SAR)地物分类方法。首先，利用像素间的空间关系和极化信息，将基于相干矩阵 Wishart距离的相似性测度和基于极化特征矢量欧氏距离的相似性测度相结合，引入马尔可夫随机场势函数，构造谱聚类的相似性矩阵；然后，采用基于 Nystr?m逼近的谱聚类实现极化 SAR数据的单次谱分类；最后，采用集成策略完成对极化 SAR图像的地物分类。实验结果表明，该算法提高了分类精度，区域一致性保持较好，且分类结果稳定。
위료강저보취류산법척도삼수대분류결과적영향，피면 Nystr?m핍근도치적분류결과불은정，제출료일충기우보취류집성적겁화합성공경뢰체(SAR)지물분류방법。수선，이용상소간적공간관계화겁화신식，장기우상간구진 Wishart거리적상사성측도화기우겁화특정시량구씨거리적상사성측도상결합，인입마이가부수궤장세함수，구조보취류적상사성구진；연후，채용기우 Nystr?m핍근적보취류실현겁화 SAR수거적단차보분류；최후，채용집성책략완성대겁화 SAR도상적지물분류。실험결과표명，해산법제고료분류정도，구역일치성보지교호，차분류결과은정。
In order to improve the robustness of spectral clustering to the scaling parameter and avoid the instable results caused by the Nystr?m approximation,a novel spectral clustering ensemble method for Polarimetric SAR (PolSAR) land cover classification is proposed. Firstly, Wishart-derived distance measure and polarimetric similarity are combined to obtain the complementary information from the spatial and polarimetric relations between pairwise pixels.The Markov Random Field (MRF) potential function is introduced to construct the similarity matrix.Then the Nystr?m approximation based spectral clustering is employed to achieve a single spectral classification of PolSAR data.Finally,multiple individual classifications are obtained and integrated by an ensemble strategy. Experimental results demonstrate that the proposed method improves the classification performance and region harmony,and leads to stable results.