Abstract: Taking surface image of iphone4S for example, for the existing vertical stripe defect and interval defect inspecting method is insufficient for high-frequency noise, a surface defect inspecting method based on projection analysis and algorithm is proposed. In order to realize automatic inspection, it designs a judgment method of vertical stripe defect based on peak valley values, and algorithms of edge line detection based on block projection and interval defect determination based integral value by analyzing the characteristics of defects. A large number of experiments show that the algorithm can let the accuracy rate above 90% in vertical and interval defect inspecting.

Key words: vertical stripe defect, interval defect, projection, high-frequency noise, peak valley value

摘要： 以iphone4S工件为例，针对工件表面竖纹缺陷和注塑缺陷检测过程中受高频噪声影响的不足，提出了基于投影分析的缺陷检测算法。通过分析缺陷的特征，设计了峰谷值判定的竖纹缺陷检测算法、基于分块投影的工件边缘线检测和注塑缺陷积分值判定算法，以实现计算机对工件表面竖纹和注塑缺陷的自动检测。通过大量实验表明：以新算法为核心技术的检测算法，能够对竖纹缺陷和注塑缺陷检测准确率达到95%以上。

关键词: 竖纹缺陷, 注塑缺陷, 投影, 高频噪声, 峰谷值