数显仪表识别在智能工厂中的整体方案设计与实现

Design and Implementation of Digital Display Instrument Recognition in Intelligent Factory

智能工厂少人化趋势逐步明显,但目前大多数生产车间的仪表识别是通过操作人员在现场采用扫描仪读取现场仪表并人工录入数据的方式。为减少人为错误和降低人工成本,强烈希望能够通过现场画面获取仪表数据或状态,并对采集到的数据进行智能识别。本文提出应用于智能工厂的数显仪表识别整体方案,通过区域定位、改进后的投影法数字分割算法以及SVM支持向量机机器学习分类识别算法对工厂仪表进行智能识别。经过大量的实验证明,该方案识别精度可达90%以上,远超过人工操作识别,是工厂智能化的发展中不可或缺的一部分,并可推广到其他类型仪表的智能识别系统。

Intelligent factory less people trend gradually obvious, at present, most of the production workshop instrument identification is through the operator in thefield using the scanner to read the field instrument and manually enter the data. In order to reduce human error and reduce labor costs, it is strongly desired to acquire instrument data or status from the live screen and to intelligently identify the collected data. In this paper, the digital display instrument identification scheme is applied to the intelligent factory. The intelligent identification of the factory instrument is realized by the regional positioning, the improved projection digital segmentation and the SVM machine learning classification recognition algorithm, after a large number of experiments show that the program identification accuracy of 90% or more, far more than manual operation identification, is an integral part of the development of intelligent factories, and other types of intelligent identification of instruments have a certain reference and technical accumulation, and can be extended to other types of instrumentation intelligent identification system.