构建了工业自动化生产设备表面缺陷检测框架,包括机械结构、照明设备、成像设备及处理算法、控制卡及软件集成模块,分析了框架各模块的作用和相互联系,面向生产线检测任务,建立系统化思维框架。以电路板表面缺陷检测应用为实例,开展了具体的缺陷检测实验,针对咖啡色电路板线路缺陷检测问题,利用局部阈值增强的二值化图像算法完成了缺陷区域检测,并实现了软件并行加速优化。实践结果表明,工程实践教学模式能够激发学生的学习兴趣,加深理论知识理解,提高动手实践能力,取得了良好的教学效果。
A framework of surface defect detection by using industrial automation production equipment is constructed. It includes mechanical structure, lighting equipment, imaging equipment and processing algorithm, and control card and software integration module. The function and the relationship of these modules are analyzed tor testing production line. The constrution process helps students to establish a systematic framework of consideration. The circuit hoard surface defec detection is taken as an example, a series of specific defect detection experiments are carried out. Aiming at the problem of brown PCB line defect detection, the image is binarized based on local threshold, the defect area detection is completed, furlhermore, the soflware parallel speed optimization is achieved. Practice results show that the engineering practice teaching mode proposed can stimulate students'interest of learning, deepen the understanding of theoretical knowledge, and improve the ability of hands-on practice. The teaching effect is satisfactory.