摘要
根据遥感历史数据,获取图像中相对稳定的边界及其特征,基于云理论,对相对稳定的各类边界建立边界云,作为标准边界特征表达并用于边缘检测。通过云生成算法对待检测边界建立边界云,建立云匹配算法,将待检测边界的边界云与标准边界云进行匹配,从而确定新的边界和边界类型,达到边缘检测的目的,实验验证了该方法的有效性。
The steady edge and its digital characteristics in image can be achieved using the RS historical data. Based on the Cloud theory, boundary clouds of all kinds of steady edge can be constructed. The boundary clouds can represent the standard edges and be used to edge detection. First, the new boundary clouds of RS image can be constructed by cloud generating algorithm. Secondly, edge cloud matching algorithm is proposed. Thirdly, matching of boundary cloud of standard edge and new edge can be accomplished and the typed of edge can be detected. The experimental results validate the efficiency of the method.
出处
《重庆邮电大学学报(自然科学版)》
2008年第6期733-736,共4页
Journal of Chongqing University of Posts and Telecommunications(Natural Science Edition)
基金
重庆邮电大学博士科研启动金资助(A2008-15)
关键词
云理论
遥感影像
空间不确定面
对象云
cloud theory
RS image
spatial uncertain area
object-cloud