In this paper, a method of tempered glass insulator recognition based on intensity in HSI modal from airborne image by helicopter on power grid is proposed. Morphology, extraction of connected components and threshold value segmentation by mean are applied to this method. The intensity of image contains the important information of insulator image, which is the base of insulator image segmentation. According to the feature of intensity in image, it is helpful that intensity image become binary image. Because there are some noises in the binary image, it is necessary that the binary image should be preprocessed to reduce noise. The erode process in morphology is used here. After above preprocessing, extraction of connected components can mark every 1 area in binary image, which includes the part of insulator and noise. Then setting threshold value range with mean in the number of every connected component is the threshold in classifying these components to insulator or not. Experimental results indicate that the method can classify insulator in image with little mistake.
[1]
Xingliang Jiang,et al.
Comparison of DC Pollution Flashover Performances of Various Types of Porcelain, Glass, and Composite Insulators
,
2008,
IEEE Transactions on Power Delivery.
[2]
Wuming Zhang,et al.
Automatic Extraction of Power Lines from Aerial Images of Unmanned Aerial Vehicles
,
2007,
IEEE Geoscience and Remote Sensing Letters.
[3]
Jia-Ping Wang,et al.
Stochastic Relaxation on Partitions With Connected Components and Its Application to Image Segmentation
,
1998,
IEEE Trans. Pattern Anal. Mach. Intell..
[4]
An Jubai,et al.
A Method of LVQ Network to Detect Vehicle Based on Morphology
,
2009,
2009 WRI Global Congress on Intelligent Systems.
[5]
Yao Degui,et al.
Inverse application of charge simulation method in detecting faulty ceramic insulators and processing influence from tower
,
2006,
IEEE Transactions on Magnetics.