论文部分内容阅读
由于指针式仪表人工检测时存在效率低、精度低、主观因素影响大、浪费大量劳动力等种种弊端。针对这种情况,提出基于机器视觉技术与自适应模板匹配技术相结合的指针式仪表不同旋角的自动检测系统.该系统首先通过千兆以太网卡和CMOS采集图像,然后基于机器视觉技术对图像进行学习匹配模板,ROI提取和阈值分割,最后通过调用由VC++编写的指针刻度识别算法,实现指针式仪表刻度的自动检测。本系统内部通过对模板的自适应匹配,获得最佳数据检测结果,从而避免了传统检测系统对仪表左旋、右旋等微小偏转的敏感性,引领系统向着高鲁棒性、高智能化方向发展。
Due to the low efficiency, low accuracy, subjective factors, waste of a large amount of labor force and other disadvantages due to the manual detection of the pointer instrument. In view of this situation, this paper proposes an automatic detection system based on the combination of machine vision technology and adaptive template matching technology to detect the different angle of rotation of the pointer instrument.Firstly, the system collects images through Gigabit Ethernet and CMOS, and then based on machine vision technology, Learning matching template, ROI extraction and threshold segmentation, and finally by calling VC ++ prepared pointer calibration algorithm to achieve automatic detection of pointer instrument calibration. The system internally matches the template adaptively to obtain the best data detection result, thus avoiding the sensitivity of the traditional detection system to small deflection such as left-handed rotation and right-handed rotation of the instrument, leading the system to develop towards high robustness and high intelligence .