Study On Hydrophobicity Of Composite Insulators Based On Canny Operator And Neural Network

With the development of our society,the construction of the power system is more and more perfect.The insulator,as an important equipment of the power system,is developing constantly.But with the running time,composite insulators occur in different degrees of aging under the external environment force,the hydrophobicity is getting worse,prone to accidents.At present,there is not a set of automatic testing system for the hydrophobicity of composite insulators.This paper analyzes several methods of composite insulators hydrophobicity detection,and summarized,based on the hydrophobicity of digital image processing,to study the role of Canny operator and BP in the hydrophobic classification field.(1)The algorithm of the minimum of RGB color component in gray image,which turn the color composite insulators hydrophobic image into gray image,and enhance the difference of gray value between the background and the water(water mark).The water(water mark)edges more clearly.(2)The improved algorithm,multi parameter H adaptive smoothing gray image,can reduce the original universal adaptive filtering due to the hydrophobicity of composite insulator image's difference,so the application scope of the algorithm will be more widely.(3)Canny algorithm is proposed based on the gradient image to cut the gray image,the algorithm to get the two value image edge is continuous,and using morphological method modified two value image,extracting two value image feature values by BP neural network to identify the level of hydrophobicity of composite insulators.In this paper,the hydrophobicity of composite insulators is simulated and analyzed on the Visual Studio 2010 software platform.The simulation experiments are carried out to improve the performance of these algorithms,which can satisfy the demand of the hydrophobicity of composite insulators.