摘要
提出一种基于标记矫正的目标特征提取算法。在目标标记过程中,使用尽量少的标记号标识二值像,同时统计各个标记的面积、灰度以及其他特征信息,并将相邻标记写入邻接表。在对邻接表整合后,将相邻标记的特征信息合并,从而获得视场中各目标的特征。仿真结果表明,该算法具有很好的效果。
A target feature detecting algorithm based on label rectification is presented. The binary image is labelled with less label number, and the area,grey and other features of each label are accounted, and the adjoining label is written into adjoining label（AL） table. After conformed the adjoining label table, the feature information of adjoining label is merged to get the target features. The simulation result shows that this algorithm has satisfactory effect.
出处
《航空兵器》
2006年第6期34-37,共4页
Aero Weaponry
关键词
标记矫正
二值像
邻接表
标记合并
特征
label rectification
binary image
adjoining label table
label merging
feature
