基于Hough变换和HSV彩色空间的电气盘柜状态识别

Status Recognition of Electrical Cabinet Based on Hough Transform and HSV Color Space

摘要:

电气盘柜的状态由安装在盘柜上的指示灯的颜色和亮灭状态表征。本文提出了一种基于Hough变换和HSV彩色空间的电气盘柜状态智能识别方法。该方法首先对实时采集的盘柜图像进行预处理,得到二值化的边缘轮廓图像；再利用霍夫梯度法检测出二值化边缘轮廓图像中包含指示灯的圆,并通过RGB阈值筛选、连通域标记法和圆轮廓阈值筛选条件对无效圆进行剔除处理；然后根据这些圆圆心处图像的HSV特征值的V值判断指示灯的亮灭状态；最后根据圆边缘处图像的H、S、V值建立点亮状态的指示灯颜色判别模型,智能判断点亮指示灯的颜色类别(红、绿、黄)。实验表明,该方法能较快速、准确地识别出处于点亮状态的指示灯及其颜色,准确度达到98%以上。

Abstract:

The state of the electrical cabinet is characterized by the color and on and off state of the indicator light installed on the cabinet. A method based on Hough transform and HSV color space is proposed and used to recognize the state of electrical cabinet intelligently in this paper. Firstly, the method carries out the pretreatment for the cabinet image captured in real time, and gains the binary edge image. Secondly, the Hough gradient method is used to detect the circle containing the indicator light in the binary edge image, and the invalid circle is removed by RGB threshold screening, connected domain labeling and circle contour threshold screening. Then, according to the V value of the HSV eigenvalues of the images at the center of these circles, the on and off state of the indicator light is judged. Finally, based on the H, S and V values of the image at the edge of the circle, the color discrimination model of the indicator light is established, and the color categories (red, green and yellow) of the indicator light are intelligently judged. Experimental results show that this method can identify the indicator light and its color quickly and accurately, and the accuracy is above 98%.