基于图像处理的陶瓷产品缺陷识别研究  

Study on the recognition method of ceramic products defects based on image processing

作　　者：李业富[1] 赵玉刚[1] 姜文革[1] 

机构地区：[1]山东理工大学机械工程学院,淄博255049

出　　处：《现代制造工程》2014年第5期109-112,共4页Modern Manufacturing Engineering

摘　　要：针对生产线上陶瓷天线罩缺陷(裂纹缺陷和气孔缺陷)图像的特点,研究了基于图像处理的天线罩缺陷识别算法。图像经灰度化处理、二值化处理、中值滤波、边界提取和目标区域标记等预处理方法,有效地去除了干扰噪声,并突显出缺陷区域。裂纹缺陷经骨架提取处理和长度统计,得到裂纹长度值;统计气孔缺陷区占有像素数,得到气孔面积值。实验表明该算法能够较精确地检测出陶瓷天线罩的缺陷值,基于图像处理的天线罩缺陷识别方法可行、有效。According to the characteristics of ceramic radome defects （crack defect and porosity defect） image in production line, the radome defect recognition algorithm was researched based on the image processing. The noise interference was effectively re- moved after image preprocessing such as gray processing, binary image processing, median filtering, boundary extraction and mark the target objects, and the defect zone was highlighted. The crack defect will be recognized after extraction backbone processing and length calculation, and the porosity defect will be recognized after area calculation. Experiments show that the algorithm can detect ceramic radome defects value accurately, so the radome defect recognition based on image processing method is feasible.

关 键 词：图像处理 缺陷识别 长度统计 裂纹检测 

分 类 号：TP391.4[自动化与计算机技术—计算机应用技术]