摘要
为了解决零件离线人工检测效率低、精度低的问题,设计了包括图像采集模块、图像处理识别模块、零件传送筛选模块的基于机器视觉的汽车零件自动筛选系统,提供了传感器触发CCD捕捉图像的信号调理电路,研究了适合汽车零件筛选的数字图像预处理、分割及模式识别方法。测试结果表明该系统运行平稳、检测迅速、分类准确,从而降低人力成本,提高了零件检测的生产效率、产品质量和自动化程度,该项目还可以推广到其他表面质量检测的行业中。
For the purpose of solving the problem that off-line manual detection is of low efficiency and not accurate enough, the automatic automobile parts recognition and classification system based on machine vision was designed, which included image collection module, image processing and recognition module and the delivering and classification module. The paper proposed the regulating circuit for trigger signal to CCD from sensor and researched the methods for sorting automobile parts, including digital image preprocessing, segmentation and pattern recognition. Test results demonstrate that the whole system operates smoothly, in- spects fast, classifies precisely. Thus it improves the production efficiency and product quality, increases the degree of automation and reduces personnel costs. At the same time, the machine vision detecting way can be expanded to other applications besides au- tomobile parts.
出处
《仪表技术与传感器》
CSCD
北大核心
2009年第9期97-100,共4页
Instrument Technique and Sensor
基金
国家自然科学基金(60374047)
浙江省科技计划重点项目(2006C23047)