Segmentation Algorithm Of Hydrophobic Image Based On Snake Model

Methods of processing insulators hydrophobicity which take the active contour model as the major content are researched in this paper.Based on deformable model, three kinds of hydrophobic processing methods are proposed and examined by experiments.A new method based on Balloon-Snake model,which confirm the energy function and the relevant weighting parameters by the prior knowledge of hydrophobic image,is efficient to extract shape information of droplets.The algorithm based on the combination of Region growing algorithm and C-V model can solve the problem of initial contour.It does not depend on the size of the image but the presegmentation region numbers,which can largely simplify the iterative calculation.S-L model is introduced based on the merits of both the Snake model and the Level set model.The extraction based on the S-L model is improved by the adaptive energy item,solving the problems of the original Snake model.The proposed approach results in improved extraction and provides evidences for contamination grades recognition of insulators.