基于优化背景差分法光学薄膜表面检测系统

Detection System of Optical Thin Film Surface Based on Optimized Background Difference Method

切膜机生产线上,需将膜切成各种规格的产品。但是在具体的操作过程中,其表面将产生诸多类型的缺陷,因此必须加以检测,例如,污渍、划痕等问题的检测。本文根据过去人工法存在效率相对较低的现实问题,阐明基于计算机视觉的自动检测设计方案,并且阐明优化的背景差法,这种方法对检测非常适用。经过验证,该技术非常有效、算法可行,能够快速检测其中存在的缺陷,在很大程度上改善了生产率。在明显缩短剪切时间的基础上,降低了生产原料的损失,极大地迎合市场需求。同时通过新的检测方法,以及均值滤波得到目标的背景,研究发现,其可以明显改善传统方法的不足。

The film cutting machine production line needs to cut the film into various specifications. However, in the process of operation, the surface has many types of defects, therefore, it must be detected, such as the detection of stains, scratches and other issues. In this paper, according to the practical problems of the past manual method, the paper expounds the design scheme of automatic detection based on computer vision, and expounds the background difference method, which is very suitable for the detection. After verification, we can find that the technology is very effective and feasible, and it can quickly detect the defects, and improve the productivity to a large extent. On the basis of shortening the shearing time obviously, the loss of raw material is reduced and the market demand is greatly met. At the same time, through the new detection method and the mean filter to get the target background, the research finds that it can significantly improve the shortcomings of the traditional methods.