基于HALCON的注射器针头缺陷检测方法

Defect Detection Method of Syringe Needle Based on HALCON

针对当前大多数医疗器械企业对注射器针头的缺陷检测仍采用人工目测的方法检测的问题,提出了基于HALCON的注射器针头缺陷检测方法。首先对被测针头在预处理阶段通过空、频域处理拉开前景与背景的对比度,然后通过模板匹配实现倒装针头的检测。再对被测针头通过Blob分析、形态学运算得到边缘信息并拟合边缘曲线,最后设计识别算法实现对弯钩针头的检测。实验结果表明:针头误检率为5%,倒装针头检测率100%,弯钩针头检测率100%,满足企业实际生产的需求。

Aiming at the problem that most medical equipment enterprises still adopt the method of manual visual inspection to detect the defects of syringes needles,a method for detecting needle defects in syringes based on HALCON is proposed.Firstly,the contrast between foreground and background is drawn through space and frequency domain processing in the pretreatment stage,and then the detection of inverted needle is realized through template matching.Then the edge information were obtained by blob analysis and morphological operation,andthe edge curve was fi tted.Finally,the recognition algorithm was designed to realize the detection of the crooked hook head.The experiment proved that the needle error detection rate was 5%,the inverted needle detection rate was 100%,and the crooked hook head detection rate was 100%.It meets the actual production needs of enterprises.