基于分形理论与SVM的河冰高分遥感影像智能识别方法研究

Research on intelligent recognition method of river ice remote sensing image based on fractal theory and SVM

黄河流域冬季凌汛灾害频发,抢险难度大,可能造成沿岸巨大的经济损失和严重的社会影响。本文针对黄河河冰高分遥感影像的几何特征、空间特征和纹理特征等复杂信息,提出一种高分遥感影像凌汛灾害监测信息识别提取方法。基于改进的ε-毯子分形理论开展河冰遥感影像的分形边缘检测与分割,利用支持向量机算法,智能分类识别不同种类河冰空间分布信息。本文研究方法可快速且准确地提取各类河冰纹理特征和区分河冰的类别,总体分类精度达到93%以上。本研究成果可为凌汛灾害风险的管理调控提供一定的决策依据。

Ice flood disasters in the Yellow River basin are frequent and difficult to rescue in winter,which may cause huge economic losses and serious social impact along the river.Aiming at the complex information of geometric feature,spatial feature and texture feature of high-resolution remote sensing image of Yellow River ice,this paper proposes an ice flood disaster monitoring information identification and extraction method.Based on the improvedε-blanket fractal theory,the fractal edge detection and segmentation of river ice remote sensing images are carried out,and the support vector machine algorithm is used to intelligently classify and identify different types of river ice spatial distribution information.The research method in this paper can quickly and accurately extract various river ice texture features and distinguish the types of river ice,with an overall classification accuracy of more than 93%.The results of this study can provide a decision basis for the management and regulation of ice flood disaster risk.