基于自适应遗传算法的超平面分类及遥感应用

发布时间：2018-04-11 05:10

本文选题：多光谱遥感影像 + 智能分类算法　；参考：《系统工程理论与实践》2017年03期

【摘要】：智能分类算法是遥感影像分类研究的热点,遗传算法作为一种智能全局优化技术在遥感影像分类中具有良好应用前景.针对现有多光谱遥感影像分类方法的不足,提出了基于自适应遗传算法的超平面分类方法(hyper plane-adaptive genetic algorithm,HP-AGA)并应用于遥感影像分类,该方法利用神经网络中的神经元激活函数Sigmoid函数,对遗传算法中交叉率、变异率进行非线性自适应性调整,不再需要反复训练遗传参数,同时利用快速全局寻优特点,确定分类超平面的各个位置参数,从而获取最佳分类超平面集进行分类.多光谱遥感影像分类方法的应用实验表明,基于自适应遗传算法的超平面遥感分类方法能更快、更稳定地收敛到全局最优解,具有更好的效率及鲁棒性,并能取得优于简单遗传超平面分类算法及传统分类方法的分类精度.
[Abstract]:Intelligent classification algorithm is a hot topic in remote sensing image classification. As an intelligent global optimization technology, genetic algorithm has a good application prospect in remote sensing image classification.In view of the shortcomings of the existing multispectral remote sensing image classification methods, a hyperplane plane-adaptive genetic algorithm based on adaptive genetic algorithm (AGA) is proposed and applied to the classification of remote sensing images. The neural activation function (Sigmoid) is used in this method.The crossover rate and mutation rate in genetic algorithm are adjusted by nonlinear self-adaptability, and the genetic parameters need not be trained again and again. At the same time, by using the characteristics of fast global optimization, the location parameters of the classification hyperplane are determined.In order to obtain the best classification hyperplane set for classification.The application experiments of multi-spectral remote sensing image classification method show that the hyperplane remote sensing classification method based on adaptive genetic algorithm can converge to the global optimal solution more quickly and stably, and has better efficiency and robustness.The classification accuracy is better than that of simple genetic hyperplane classification algorithm and traditional classification method.
【作者单位】：中南大学地球科学与信息物理学院中南大学空间信息技术与可持续发展研究中心;长沙市发展和改革委员会;Department
【基金】：国家自然科学基金(41171326,40771198)~~
【分类号】：TP18;TP751

【相似文献】