Abstract: This paper designs an on-line detection system for work-pieces assembling error, which solves the problems of high labor intensity and low efficiency during the error detection done by people and considers the special feature of the detected work-pieces. The key idea of the system is that it introduces the non-contact vision inspection technology to industrial production. Utilizing image processing technologies such as noise removal, edge extraction, shadow detection, parsing of pattern feature, template matching. etc, the work-pieces whose surface images are captured by the CCD camera are detected whether they have processing errors or not during the production procedure. The experimental result indicates that the designed system has the merits of reducing the operator’s labor intensity, improving the efficiency of the error detection.

Key words: image processing, online detection, parsing, character recognition

摘要： 针对加工零部件人工误差检测高劳动强度、低效率的问题，以及被检测工件的特殊形貌，设计了一套加工件装配过程在线检测系统，将基于计算机的非接触式视觉检测技术引入到工业生产线中。利用CCD相机捕捉工件表面形貌图像，结合图像噪声去除、边缘提取、阴影检测、模式特征句法分析、模板匹配等技术，在线检测工件生产流程中出现的误差。实验结果表明所设计的系统降低了操作人员的劳动强度，提高了零部件误差检测效率。

关键词: 图像处理, 在线检测, 句法分析, 字符识别