Research And Implementation Of Image Enhancement Algorithm Under Transformer Oil

With the rapid development of various industries in the country,the demand for electricity in the whole society is also increasing,limited by the distance between power plants and users,most areas of electric energy rely on power supply and distribution systems to provide.Transformer is one of the important equipment in the power supply and distribution system in our country,the insulation and cooling medium inside transformer are changing constantly with the progress of industry,therefore,the oilimmersed transformer using it as the insulation and cooling medium can be widely used.Oil-immersed transformers are the main force of power distribution in China’s power infrastructure and rural electrification projects,in order to improve the safety,economy and reliability of their operation,they should be regularly detected to extend the life of equipment and reduce the incidence of accidents.For the sake of safety and high efficiency,many scholars at home and abroad have designed inspection robots that can be applied to the oil-immersed transformer,and through the analysis of the image to locate the fault point.However,the transformer oil will deteriorate and change color with the increase of service life,so it is necessary to process the image clearly in order to locate the internal fault accurately.In this paper,a variety of image processing algorithms applied in different environments are studied in depth,and an image enhancement algorithm for transformer oil environment is proposed,which aims at improving image quality,to improve the accuracy of fault early warning.The specific study is as follows:First of all,the image is prepossessed by color correction and contrast enhancement.Then,the results of the two preprocessing are fused by the method of image pyramid to get the transformer oil image with good display effect but with noise.Finally,the image is finally sharpened by the traditional denoising method and the resolution enhancement method based on deep learning.The experimental results show that,compared with the original image,the oilimmersed transformer image processed by this algorithm has better contrast effect and color display,the two-dimensional entropy of the image and the value of underwater color image quality evaluation are increased by 23.56% and 22.04% respectively,the number of feature points matching is increased by 8.05 times,and the gray distribution in color histogram experiment is better.The experimental results prove that the proposed algorithm can effectively improve the image quality of oil-immersed transformer and provide a basis for internal fault detection of oil-immersed transformer.