The paper presents a novel region-based active contour model designed to address the challenges posed by intensity inhomogeneities in images. The model incorporates a region-scalable fitting (RSF) energy, which uses intensity information from local regions at a controllable scale to guide the contour evolution. The RSF energy is defined using a kernel function that allows for the selection of the scale parameter, enabling the model to handle intensity variations from small neighborhoods to the entire image domain. The energy is formulated in a variational level set framework, incorporating a level set regularization term to ensure the regularity of the level set function and avoid reinitialization. The proposed method is demonstrated to be effective in segmenting images with intensity inhomogeneities, as shown through synthetic and real images. Experimental results highlight the model's robustness to noise and its ability to accurately segment objects with weak boundaries. comparisons with other methods, such as piecewise smooth models and mean shift algorithms, further demonstrate the优越性 of the proposed method in terms of computational efficiency and accuracy. The paper also discusses the influence of the scale parameter on segmentation results and suggests extensions to multiphase level set formulations and narrow band implementations to enhance performance.The paper presents a novel region-based active contour model designed to address the challenges posed by intensity inhomogeneities in images. The model incorporates a region-scalable fitting (RSF) energy, which uses intensity information from local regions at a controllable scale to guide the contour evolution. The RSF energy is defined using a kernel function that allows for the selection of the scale parameter, enabling the model to handle intensity variations from small neighborhoods to the entire image domain. The energy is formulated in a variational level set framework, incorporating a level set regularization term to ensure the regularity of the level set function and avoid reinitialization. The proposed method is demonstrated to be effective in segmenting images with intensity inhomogeneities, as shown through synthetic and real images. Experimental results highlight the model's robustness to noise and its ability to accurately segment objects with weak boundaries. comparisons with other methods, such as piecewise smooth models and mean shift algorithms, further demonstrate the优越性 of the proposed method in terms of computational efficiency and accuracy. The paper also discusses the influence of the scale parameter on segmentation results and suggests extensions to multiphase level set formulations and narrow band implementations to enhance performance.