摘要
在图像模式识别中 ,面积是很重要的特征。提出一种计算面积的新方法 ,利用Freeman链码矢量分析 ,对边界像素标注综合处理 ,再进行边界像素坐标加权求和计算 ,求得目标面积 ;该方法与原有计算面积的方法相比 ,该方法实现简单 ,计算量小 。
In the image recognition, the object's area is a very important characteristic. In this paper, a new and easy realized algorithm for calculating the area is presented. Freeman boundary encode vector is used for analysis. The edge pixels'label is processed Comprehensively. Then, add the edge pixels' coordinate to coefficient we get the object's area. Compared with the old algorithm, its required operations are reduced and its result is good.
出处
《国防科技大学学报》
EI
CAS
CSCD
北大核心
2002年第4期61-64,共4页
Journal of National University of Defense Technology