基于OTSU-Graham改进算法的保护压板状态辨识研究

随着自动化与智能化技术在变电站中的推广应用为智能巡检奠定了基础，但是目前二次设备的保护硬压板仍然大多采用人工现场核对的方式，存在核对频次低、校对过程溯源性不足的问题。为此，本文提出了一种基于图像内容识别的压板运行状态智能识别方法。首先采用基于空间领域信息的OTSU算法进行阈值分割消除光照不均阴影区域的影响，在此基础上基于Graham的最小外接矩形算法检测压板开关的最小矩形面积，通过识别面积大小来判断压板是否投入。该方法能够有效减少阴影干扰的影响，准确辨识图像中压板的运行状态。

A Study on the Identification of Protection Press Plate State Based on OTSU-Graham Improved Algorithm

With the popularization and application of automation and intelligence technology in substation, the foundation of intelligent inspection is laid, but at present, the protection hard-pressed plate of secondary equipment still mostly adopts manual on-site checking, which has the problems of low checking frequency and insufficient traceability of proofreading process. Therefore, this paper proposes an intelligent recognition method of the operating state of the press plate based on image content recognition. The OTSU algorithm based on spatial domain information is first used to segment the threshold to eliminate the influence of illumination uneven shadow area. On this basis, the minimum rectangular area of the plate switch is detected based on the Graham minimum external rectangular algorithm, and the area size is identified .