基于模糊双支持向量机的遥感图像分类研究

遥感图像的分类是研究土地利用变化的基础。传统的遥感图像分类方法存在运算速度慢、精度比较低和难以收敛等问题。提出了一种基于模糊双支持向量机的多类分类方法,将模糊技术引入到双支持向量机中,赋予不同样本以不同的模糊隶属度,然后将模糊双支持向量机推广到多类分类中,最后将新方法应用到遥感图像分类中。实验表明,新方法比传统的支持向量机多类分类方法有较高的分类精度,并且有较强的抗噪声能力,在运行时间上也是可行的。模糊双支持向量机是一种有效的遥感图像分类方法。

Research of Remote Sensing Image Classification based on Fuzzy Twin Support Vector Machine

The classification of remote sensing image is the basis of studying the land change.Some traditional algorithms of remote sensing image classification have some problems such as low computing rate,low accuracy and hard for convergence.A multi-classification algorithm based on fuzzy twin support vector machine is presented.First,fuzziness is inducted into twin support vector machine by applying the fuzzy membership to every training sample,then fuzzy twin support vector machine is extended to multi-classification,finally,remote sensing image classification is achieved by the proposed method.The study indicate the proposed method is more precise than traditional support vector machine multi-classification algorithm,and has high anti-noise performance and feasible runtime.Fuzzy twin support vector machine is an effective method of remote sensing image classification.