Research On Recognition Method Of Composite Insulators Hydrophobicity Based On Image Processing Technology

The hydrophobic condition of insulator is related to the safe operation of theinsulator. In order to recognize the surface hydrophobic grades of insulator conveniently,quickly and accurately, a new method of measuring hydrophobicity level based on imageprocessing technology is presented. Firstly, One hundred and forty insulators which aredifferent operating life, hydrophobicity,pollution level,manufacturer are selected, andthese insulators are used to do spraying grade and natural rain experiment, at the sametime, taking photos of these insulators and getting a lot of image samples. Secondly,thehomomorphic filter and adaptive histogram equalization are used to enhance thehydrophobic image. Then the improved Canny algorithm is used to extract the edge ofwater (or water mark), and the mathematical morphology is used to amend it, getting thefinal segmented image. Four features associated with hydrophobicity are extracted in theimage, which are as the input vector and the hydrophobic level are as output vector, thehydrophobic recognition model based on BP neural network is founded; Finally, in theMATLAB software platform, using the algorithm proposed in this paper as the basis, aset of composite insulator hydrophopic automatic analysis was developed. Thehydrophobic grades of the test samples are recognized by the software. Experimentsshow that the hydrohopic software can overcome the influence of human factors,effectively recognize the seven typical hydrophobic grades of insulator. The totalrecognition rate is90%and it is an effective recognition method for online detection ofinsulator hydrophobicity.