Characterization of the algae growth on the RTV-coated insulator surface based on the corrosion expansion and the image segmentation algorithm

Image recognition technology for RTV (room temperature vulcanized)-coated silicone rubber insulators with algae fouling is gradually gaining popularity due to time-consuming and laborious nature of manual inspection. However, there is still an issue with the accuracy of image recognition. In this paper, a corrosion and expansion algorithm is applied for digital image processing to accurately characterize algae growth on silicone rubber insulator surfaces. After pre-processing, two common indexes are used: EXG (excess green index) and HVI (hague vegetation index). The effect of image segmentation is studied, and then the characteristic green value is used to measure algae greenness. The method for measuring algae cell density is also described. By calculating the characteristic green value of collected insulator images, the corresponding algae cell density is determined, and the algae growth degree is characterized along with the algae coverage rate. Therefore, the erosion and dilation algorithm is an essential part of image pre-processing. Based on the above work, the relationship curve between the characteristic green value and algae cell density is accurately obtained. The research results are of great significance for characterization of algae pollution on insulator surfaces using image methods.