A Variational Level Set Image Segmentation Method via Fractional Differentiation

To solve the issues with conventional level set segmentation algorithms, which are sensitive to initial contours and less noise-resistant, a model based on coupling of texture information structural is developed. In this model, rotation invariant mask produced by fractional-order differentiation used first describe image’s global information. Then, power function energy generalization solved applying factorization theory, for each pixel image, not only its but also surrounding taken into account integrated via weight scaling. At same time, L2 norm image difference from fitted generate model. The final results study demonstrate that proposed achieved better performance than current active contour models in terms robustness Gaussian noise pretzel noise, as well accuracy algorithm running time. These were obtained synthetic images, real natural images.