广义LVQ算法及其在遥感影像分类中的应用研究

The Remote Sensing Image Classification Based on Generalized Learning Vector Quantization Algorithm

在比较了学习矢量量化(LVQ)算法和广义学习矢量量化(GLVQ)算法的基础上,建立了基于GLVQ的遥感影像分类模型。以实际土地覆盖分类为例,通过与传统统计方法和LVQ分类器比较,GLVQ分类器具有分类正确率高,收敛速度快,适应范围广等优点。

After comparing Leaming Vector Quantization （LVQ） algorithm and Generalized Learning Vector Quantization （GLVQ） algorithm, this paper establishes a GLVQ-based classification model for remote sensing image. With the experimental applications of land-over classification with the presented model, the GLVQ classifier has higher recognition rate, faster convergence speed and wider adaption range than conventional classifier and LVQ classifier.