基于人工免疫系统的带钢表面缺陷检测技术  

Surface defects detection technique for steel strips based on artificial immune system

作　　者：许小润[1] 吴贵芳[1] 

机构地区：[1]河南科技大学电子信息工程学院,河南洛阳471003

出　　处：《计算机应用》2010年第8期2247-2249,2253,共4页journal of Computer Applications

基　　金：国家自然科学基金资助项目(50074010)

摘　　要：针对传统的图像检测算法对具有噪声以及光照不足或不均匀的带钢表面图像检测时效果不理想的问题,提出了基于人工免疫系统的带钢表面缺陷检测技术。根据检测器和抗原之间的论域空间包含关系,以及自体在论域空间中的位置信息,引进了分块空间的方式,提出了检测器分块的生成算法,对缺陷图像表面灰度信息进行三维恢复,获得带钢表面缺陷的高度信息,实现其三维质量检测。实验表明,该方法无论是在对比度低、光照不均还是有噪声干扰的情况下相对于传统的图像检测算法具有较好的处理效果。The traditional algorithms of surface defects detection for steel strips cannot achieve expected results because of the image noises and inadequate or uneven illumination.To address this issue a new method based on artificial immune system was proposed.According to the spatial inclusion relationship between detector and antigen as well as self location information in domain block-based detector generating algorithm was introduced to obtain the height information of surface defects for steel strips through three-dimensional recovery of their gray-level images thus the three-dimensional quality of surface defects detection for steel strips was achieved.The experimental results show that the proposed method has an excellent effect compared with the traditional image detection algorithm in the condition of low contrast ratio uneven illumination or noise interference.

关 键 词：带钢 表面缺陷 人工免疫系统 检测器分块 三维恢复 

分 类 号：TP317.4[自动化与计算机技术—计算机软件与理论] TP311[自动化与计算机技术—计算机科学与技术]