支持向量机在空间信息处理领域的应用研究

The application and research progress of SVM in spatial information processing

支持向量机作为一种最新的也是最有效的统计学习方法,近年来成为模式识别与机器学习领域一个新的研究热点。支持向量机具有小样本学习、抗噪声性能好、学习效率高和推广性好的优点,能够用于空间信息处理分析领域的遥感影像处理、高光谱分类、拟合与回归、数据挖掘、目标检测等任务。本文在总结分析近年来支持向量机在空间信息处理领域应用主要进展与成果的基础上,结合支持向量机理论方法与空间信息处理的发展趋势,提出了今后有必要重点研究的若干问题,包括空间数据挖掘、智能空间信息处理、高维空间数据处理等。

As one of the most popular and effective statistical learning algorithm,Support Vector Machine(SVM)has become a new hot topic in pattern recognition and machine learning fields in recent years.SVM has some advantages,for example,its applicability to limited samples,robustness to noises,high learning efficiency and easy generalization,and those are commonly difficult issues in spatial information processing.So SVM can be used to spatial information processing,and some major applications include RS image processing,hyper-spectral RS data classification,spatial approximation and regression,data mining,target identification,and so on.Based on a brief introduction and summary to the achievements of SVM used to spatial information processing,some potential applications,including spatial data mining,intelligent spatial information process and high-dimensional spatial data processing are discussed in detail.