探讨了利用机器视觉技术进行注射模具型腔异常检测的流程和关键技术，包括型腔图像预处理、型腔图像模板匹配、互相关方法，并对一工作过程中的模具进行了型腔异常检测实验。结果表明，本检测方法准确性高，具有实用价值。分析了对塑料少量残留型腔的误判原因，为进一步优化检测奠定了基础。%The procedures and key technologies including the methods for pretreatment of cavity image, templet matching of cavity image and cross-correlation of cavity abnormal situation for injection mold based on machine resion technology,and the cavity abnormal situation of an injection mold during working was detected. According to the experiment results,this detection method is of high accuracy and practical value. The reason for misjudgment of the cavity in which a little material leaves was analyzed, which lays the foundation for the further optimization of the detection method.
展开▼
