利用图像识别技术实现复杂装备的状态辨识对于装备状态实时判读、及时干预装备故障、提高装备保障能力具有重要作用。针对复杂装备元件数量多、类型多、流程复杂的特点,设计了复杂装备图像识别技术体系,采用图像识别技术分别进行识别边界判定、图像分割、图像状态识别,通过建立装备状态辨识模型,实时判断装备状态。利用OpenCV工具中的图像识别函数对某装备进行状态辨识试验,验证了技术体系的可行性及各技术的正确性。

Complex equipment statement identification based-on image recognition is important for real-time interpretation of equipment statement, timely intervention equipment faults and improving equipment support capability. According to the characteristics of large quantity of component, complex process, the paper designs a technical complex equipment identification architecture based on image recognition. The technical architecture includes: image edge judgment, image segmentation, statement identification. The testing of equipment statement identification with OpenCV tools demonstrated that the proposed technical architecture is feasible and correct.