摘要
给出一种图像标记和目标特征统计算法,该算法对分割后的图像按行操作:依次对每行中独立的块(Block)进行标记,并统计其特征值。然后对每两行的Block进行比较,判断其是否粘连。若是,则归为一类(Variety);若否,则定为新类。试验结果证明了此算法的可行性。
An statistical algorithm of image label and target characteristics is presented. With this algorithm every independent blocks in each row of segmented image is labelled, and their statistical features are stored. Then blocks in each double rows are compared, and blocks are classified into different varie- ties by the compared result. The feasibility of the method is proven in the experiment.
出处
《航空兵器》
2009年第4期38-41,共4页
Aero Weaponry
关键词
计算机图像处理
图像标记
目标特征统计
computer image processing
image label
statistical feature of target
