Research On Infrared Image Preprocessing Method Of Zero Resistance Ceramic Insulator Identification

Insulators are important insulation equipment and widely used in transmission lines,andis the key to guarantee the electrical performance. Overhead line insulators will bedeterioration due to being affected by the long-term erosion of environmental, electrical loadand mechanical,etc. In case zero resistance insulators are not detected and replaced timely,which affects the grid security, Under the bad weather conditions, it will greatly increase theprobability of insulator flashover, causing serious accidents such as insulator string, the stringor wire to the ground.The detection of zero resistance insulators is an important work of the electricity sector,but also the overhaul of the transmission line state. With the development of the moderninfrared technology, Infrared thermal imaging technology widely used in electric powersystem fault detection because of its safety, non-contact, not influenced by subjective factors.In the method of infrared detection zero resistance insulators based on infrared image, imageprocessing is very important. In this paper, it has done some research to contaminatedinsulators’infrared image processing methods.It is inevitable to merge many kinds of noise to image when the infrared thermographytaking images. The noise will cause image quality drop, the contrast is not obvious andfeatures to cover up, directly influence for the accuracy of The zero value insulators detectionprocedur. In this paper,according to the characteristic of infrared images, a new method ispresented for improved threshold for infrared image de-noising based on NeighShrink. Themethod adjust the size of the threshold by the different sub-bands scales, and it has goodadaptability. Through the experiment, the method is superior to other wavelet thresholddenoising method for PSNR and have a good denoising effect for the big noise and low SNRinfrared thermal image of insulator, while maintaining the integrity of the image information.In the order to sake of the zero resistance insulator string and the background separatedfrom the whole image. According to the characteristics of the zero resistance insulatorinfrared thermal image, image segmentation method on the combination of Otsu doublethresholds segmentation and a modified differential evolution is proposed. This method takesOtsu for fitness function,ensures the population diversity by variation,crossover,selectionoperation and avoids getting into the local best,and the crossover probability factor isreconstructe adaptively according to the iteration number to enhance the convergence speedand the search capability. The experimental results show that the method has a faster convergence speed and a better image segmentation effect, so as to realize the diagnosis ofzero resistance insulator.The algorithms mentioned above carried out successfully in practice application, itimproves the accuracy of image analysis and settles a basement for diagnosis of zeroresistance insulator.