<p>This paper presents a fractional Laplacian-enhanced local-global intensity fitting energy model for segmenting noisy and intensity inhomogeneous images. The proposed variational model is built on two key observations: the difference map between the Gaussian-filtered source image and its fractional Laplacian image preserves the global structure of the target object, while their addition map enhances the intensity contrast without over-smoothing. The total energy functional of the model comprises a global fitting energy of the difference map, a local fitting energy of the addition map, and a length regularization term for the segmentation contour. In minimizing the energy functional, the global fitting term drives the segmentation contour to move towards object boundaries; the local fitting term generates a local force to attract the moving contour to stop at the expected object boundary; the length regularization ensures the contour can tightly wrap the target object. We employ an iterative convolution-thresholding method to implement the model, ensuring energy decay at each iteration. Numerical experiments demonstrate its effectiveness and efficiency in addressing challenging segmentation tasks.</p>