注射制品表面缺陷在线检测与自动识别

基于机器视觉技术的注射制品表面缺陷检测与识别可有效解决人工抽样检测问题,为克服现有缺陷识别算法对于不同的制品需分别进行样本训练、图像质量要求高、可操作性差等难题,在采用图像处理技术对制品表面缺陷进行检测的同时,提出一种基于缺陷区域轮廓、制品轮廓、区域灰度等特征的缺陷自动识别算法,避免了大量制品图像样本的训练过程,提高了可操作性。基于该方法,开发了一套注射制品表面缺陷在线检测与识别系统,试验表明,对短射、飞边、裂纹3种常见表面缺陷的识别率为91.8%。

On-line surface defects detection and automatic recognition of injection moulding products

The automatic surface defects detection and recognition system for injection moulding products based on machine vision can effectively solve the problem that artificial sampling detections were of low automation degree and efficiency. In order to overcome that the existing defects recognition algorithms needed samples training for different prod- ucts respectively, required high quality images but had poor maneuverability, Using image processing algorithms to detect surface defects, the defects automatic recognition algo- rithms based on the features of defect region profiles, product profiles and regional gray were proposed to avoid samples training and get good maneuverability. And also an auto- matic surface defects detection and recognition system for injection moulding products based on the algorithms was developed. The experimental results showed the recognition accuracy rate for short shot, flash, crack is 91.8%.