一种高分辨率遥感影像分类的特征指数

›› 2016, Vol. 29 ›› Issue (11): 74-.

一种高分辨率遥感影像分类的特征指数

刘红彬，晁拴社，王兴

(重庆邮电大学通信与信息工程学院，重庆 400065)

出版日期:2016-11-15 发布日期:2016-11-24
作者简介:刘红彬（1993-），男，硕士研究生。研究方向：遥感影像的融合和分类。
基金资助:
重庆市2013博士后科研基金资助项目（RC201336）

A New Feature Index for classification of High Resolution Remote Sensing Images

LIU Hongbin，CHAO Shuanshe，WANG Xing

(School of Communication & Information Engineering , Chongqing University of Posts and Telecommunications, Chongqing 400065, China)

Online:2016-11-15 Published:2016-11-24

摘要/Abstract

摘要：

随着影像分辨率的提高，传统的光谱特征不能有效地描述复杂的高分辨率影像信息，从而影响高分辨率遥感影像的分类。为了弥补传统光谱方法的不足，提出了一种加权对象相关指数(WOCI)，并将其应用到基于支持向量机(SVM)的影像分类中。该指数是通过考虑具有相似性光谱的对象来构建的，可全面地描述影像的上下文结构。结果表明与仅考虑光谱特征和像素空间特征进行分类的方法相比，基于WOCI特征的分类结果有更高的精确性,且分类精度提高了7.16%。

关键词: 高分辨率遥感影像分类, WOCI, 光谱特征, 空间特征

Abstract:

With the improvement of image resolution,the traditional spectral characteristics cant effectively describe the information of image with high resolution and affect the classification of high-resolution remote sensing images.In order to make up for the inadequacy of traditional spectral method,it presents a novel spatial feature called weighted object correlative index (WOCI) and applys this index to image classification based on support vector machine (SVM).The index is created by the objects with similar spectrum and can fully describe the context structure of image.The experimental results show that the feature extracted in this paper has higher accuracy than those approaches that only consider spectral features or pixelwise spatial features and the classification accuracy improves 7.16%.

Key words: classification of very high resolution image, weighted object correlative index, spectral feature, spatial feature

中图分类号:

TP751.1

刘红彬，晁拴社，王兴. 一种高分辨率遥感影像分类的特征指数[J]. , 2016, 29(11): 74-.

LIU Hongbin，CHAO Shuanshe，WANG Xing. A New Feature Index for classification of High Resolution Remote Sensing Images[J]. , 2016, 29(11): 74-.