基于计算机视觉的电缆终端表面温升分析系统

摘要

为了分析谐波对电缆终端表面产生的热点升温,设计并实现了一个基于计算机视觉的电缆终端表面温升分析系统.为了定位发热电缆区域,提出了基于直方图分析的背景温度计算方法和形态滤波去噪方法,然后通过一个移动窗口在发热电缆终端内定位热点温升区.通过搭建实验平台获取的红外热图像进行相关的实验,表明提出的系统方案切实可行,分析结果准确,为后续的故障诊断分析打下良好的基础.%In order to analyze the impact of harmonic on the cable terminal surface temperature, a computer vision based on cable terminal surface temperature analysis system is designed and implemented. In order to locate the heating cable area, the background temperature is calculated based on histogram a-nalysis, and noise is removed by using a shape filterer. Then, a window is moved in the heating cable terminal to determine the hot spot temperature area. The experiments with the infrared images obtained by the experimental platform show that the proposed system is feasible; the performance of the proposed algorithm is equivalent to the result obtained by manual under various conditions. And it lays a good foundation for the subsequent fault diagnosis.