大口径精密光学元件表面疵病检测系统研究

Study on automatic flaw inspection system for large caliber precision optical surface

根据大口径精密光学元件表面疵病检测的工程特点,提出了一种基于优化模糊相似度算法的案例推理疵病识别法,解决了疵病图像获取中的摄像机平面移动定位、自动调焦和图像拼接等自动检测技术问题,研制了基于机器视觉的大口径精密光学元件表面疵病检测系统。实验结果表明,以上算法和技术是正确的,达到了理想的效果,可推广应用到其他材质的精密表面缺陷检测中。

The current situation of precision optical surface flaw inspection both in China and abroad was studied. According to characteristic of large aperture optical surface, a Case-Based Reasoning （CBR） method in flaw pattern recognition is presented, which is based on optimized fuzzy similarity algorithm. The technical problems in automatic inspection such as camera location, auto-focusing and image mosaic during flaw image acquiring process were solved, and a flaw inspection system was designed. Experiments show that the proposed algorithms and techniques are feasible and can be applied to other precision surface inspection applications.