摘要
近年来,随着高分一号卫星发射,高分辨率遥感影像得到广泛应用,面向对象的影像信息提取技术得到了迅速发展。在讨论面向对象信息提取的关键技术"多尺度分割技术"后,运用基于规则的方法来提取地物信息,探索基于规则分类方法的规则创建方法,最后根据混淆矩阵验证实验结果的精度。
In recent years, with the launch of first satellite with high resolution, high resolution remote sensing images are widely used and object-oriented image information extraction technology has been de- veloped rapidly. Based on the discussion of multi-scale segmentation technology as the key technologies of object-oriented information extraction, the rule-based method was used to extract feature information, and then rules creating methods based on rule classification were explored, and the accuracy of the final results was verified based on a confusion matrix.
作者
于辉
陈晓停
汪金花
YU Hui CHEN Xiao-ting WANG Jin-hua(College of Mining Engineering, North China University of Science and Technology, Tangshan Hebei 063009, Chin)
出处
《华北理工大学学报(自然科学版)》
CAS
2017年第1期6-11,共6页
Journal of North China University of Science and Technology:Natural Science Edition
基金
大学生创新性实验计划项目(x2015091)
关键词
高分一号
面向对象
特征提取
多尺度分割
the first satellite of high resolution
object-oriented
feature extraction
multi-scale segmentation