基于ALOS影像纹理特征的日光温室信息提取方法研究  

Extraction of Greenhouse Information Using Texture Features Within ALOS Image

作　　者：李洪伟[1,2] 段磊[1,3] 

机构地区：[1]兰州大学资源环境学院 [2]兰州军区68029部队 [3]解放军77100部队

出　　处：《现代农业科技》2016年第2期342-343,352,共3页Modern Agricultural Science and Technology

摘　　要：日光温室为解决我国西北部地区冬季蔬菜供应问题发挥了重要作用。本文通过计算不同方向的纹理特征,采用支持向量机(SVM)提取日光温室,研究纹理方向对信息提取精度的影响。结果表明:1纹理特征能提高分类精度,但提升幅度不大。2日光温室的最佳纹理方向为45°,总精度为93.57%,Kappa系数为0.90,且最佳纹理方向与地物的主方向大致相同。Greenhouse plays a critical role in coping with difficulties of winter shortage of vegetables in northwestern China. In this article, greenhouses were extracted on textural features by different direction, and using Support Vector Machine （SVM）, in which the effects of texture direction on classification accuracy were examined. The results showed that ：①texture feature could slightly improve classification aceuraey.②the best texture direction for extracting greenhouse was 45 degrees,overall accuracy and Kappa reached 93.57% and 0.90, and the best texture direction was the same as the main direction of the object.

关 键 词：支持向量机 基于对象影像分析 ALOS影像 日光温室 纹理特征 

分 类 号：TP753[自动化与计算机技术—检测技术与自动化装置]