Pattern recognition for counting of bounded bar steel
Pattern recognition for counting of bounded bar steel
复制标题
DOI：
10.1109/icadiwt.2011.6041396
复制
发表时间：
2011-10
期刊：
影响因子：
--
通讯作者：
Jing-zhong Wang;Chen Hao;Xiaoqing Xu
中科院分区：
文献类型：
--
作者：
Jing-zhong Wang;Chen Hao;Xiaoqing Xu
棒材成捆后的自动计数一直是一个悬而未决的问题。提出了一种新的基于准圆假设的分割方法和一种新的基于扫描的目标识别方法。这两种方法的加、减运算和逻辑运算简单，具有真实的时间特性。实验结果表明，该方法能有效地分割出聚集的条状物，提高了对不规则聚集物体的识别和计数的准确性和速度。
Automatic Counting of the bundled bar steel is unsolved. A new segmentation method based on quasi-circular assumption and a new object recognition method based on scanning are present. Owing to the simple addition, subtraction and logic operation, these two methods have real time property. The total CPU time is less than 1 second for an 640*480 pixels image, The experiments show that the new method can segment the aggregated bars effectively, and it can be used to improve the accuracy and the speed of recognizing and countering of irregular and aggregated objects.