The Hydrophobicity Detection Of Composite Insulator Based On Image Processing

Hydrophobic Classification has become the mainstream method of composite insulator online testing since its simple operation and low equipment requirement. However, it has high subjective reliance. Hydrophobic Classification with image processing which need extract watermark primarily provides basis for determining the hydrophobic level quantitatively in recent years. Hydrophobic image presenting low contrast,uneven illumination, bead lacking part of edge、having shadow、reflecting light and sticking together, dirt existing increases the difficulty in processing image, further leads to obtaining wrong parameter. It’s usually not easy to solve all the problems mentioned above with a solo method. Since these problems not isolated, low contrast, reflective and sticky of bead are focused. Results show the effect is good.According to the characteristics of the hydrophobic image, transforming the true color image into saturation component of HIS color space to enhance the contrast whose procedures is simple and effects is obvious. Moreover, it has part function of depressing uneven illumination.As for bead, its reflective area is treated as "hole" of image which can be eliminated through filling it simply. Morphology operation can spilt off beads having large area while less contact section but not for the opposite. For the latter, watershed transformation has better effect.Adding appropriate concentration of black pigment and snapping with function of "local color capability", picture got from which way can extract easily. Improved shape factor method is the commonly used hydrophobic indicator function whose rationality of parameters has been proved, but still not accurate enough. The criterion of forming process is analyzed, the two main factors including snap distance and snap angle are researched, and the fan is used as the effective selection of picture which all improved the accuracy of judging hydrophobic level of insulator.In order to using hydrophobic classification for offline detection, hydrophobic detection device including camera and light source configuration and hydrophobic detection software are designed for facilitating testing process.