基于机器视觉的硅片检测分类系统设计

等离子体增强化学气相沉积是太阳能硅片生产工艺流程中一道重要的工序。硅片经过该工序后的表面破损检测和减反膜颜色的检测主要通过人工完成，存在不稳定、高碎片率和低速率等问题，需要利用机器视觉技术，开发硅片视觉检测系统。文中根据具体的测试要求完成光学设备选型，利用“Visionpro”视觉开发软件设计硅片破损检测与颜色检测程序。实际使用结果表明，该系统破损测量的精度在0．2mm以内，图像颜色检测满足实际应用需求。

System Design of Silicon Detection Classification Based on Machine Vision

Plasma enhanced chemical vapor deposition is an important part of solar cell production processes. Silicon wafer surface defect detection and antireflection film color detection are mainly accomplished in manual, which causes the problems of instability, high scrap rate and low rate. The system uses machine vision technology to develop an inspection system for the silicon. According to the specific test requirements to complete the optical equipment selection, using the machine vision development software VisionPro,we write the program of the silicon breakage detection and color detection. The results of practical applications show that the measurement accuracy is within 0.2 mm. Image color detection can meet the need of the practical application.