Abstract: In order to quickly and accurately identify the hydrophobicity of composite insulator skirts, this paper proposes a hydrophobicity evaluation method of composite insulators based on an improved Grab Cut algorithm. On the basis of Grab Cut, an adaptive contrast enhancement algorithm is introduced to solve the problem of uneven image illumination and an adaptive smoothing filtering algorithm is used to solve the problem of image noise. A droplet of 20 μL is dropped on the surface of the skirt. The improved Grab Cut algorithm is used to identify the droplet area of corona aging samples, natural contaminated silicone rubber samples, different hydrophobic silicone rubber samples and composite insulators under actual operating conditions. The results show that the algorithm can effectively reduce the image noise caused by the contamination of the surface area of the skirt during the operation of the composite insulators, and can accurately segment the droplet images from the composite insulator background, which verifies the effectiveness and reliability of the algorithm. It can also measure the change of the droplet area of both artificially accelerated aging and natural aging samples, and realize the evaluation of hydrophobicity of composite insulators in accordance with the static contact angle method.