一种基于约束蚁群反馈算法钢坯图像分割技术  

Billet Image Segmentation Techniques Based on Feedback Constraints Ant Colony Algorithm

作　　者：乔木[1] 王天顺[1] 裴洪文[1] 

机构地区：[1]郑州升达经贸管理学院信息工程系,郑州451191

出　　处：《科技通报》2014年第2期173-175,共3页Bulletin of Science and Technology

摘　　要：结合蚁群反馈优化算法、多属性约束理论的优点,提出了一种基于约束蚁群反馈算法的钢坯图像分割算法。算法采用钢坯图像边界粗糙熵作为分割的反馈标准,充分利用蚁群算法的反馈调节能力,对属性像素进行约束,保证在约束条件下的最优化边沿分割。经过实验测试,结果表明,由于使用约束蚁群反馈算法避免了钢坯像素边沿混合弊端,采用模糊边界计算原则,完成边沿像素分割,所以不仅得到了令人满意的钢坯分割效果,而且大大的提高了计算速度,是一种实用有效的钢坯图像分割方法。This article feedback combined with ant colony optimization algorithm, the advantages of multiple attribute con-straint theory, proposed an ant colony based on constraint feedback billet image segmentation algorithm of the algorithm. Algorithm USES the billet image boundary rough entropy as the division standard of feedback, and it makes full use of the feedback adjustment ability of ant colony algorithm, constraints on properties of pixels. Results show that the feed-back due to the use of ant colony algorithm to avoid the drawbacks of billet mixed pixel edge, fuzzy boundary calculation principles, complete division of edge pixels, so not only got a satisfactory segmentation results of billet, and greatly im-proves the computing speed, is a practical and efficient billet image segmentation method.

关 键 词：钢坯图像分割 蚁群算法 约束 

分 类 号：TP391.4[自动化与计算机技术—计算机应用技术]