Coronal mass ejections (CMEs) are considerable magnetized plasma eruptions from the corona of the Sun towards the heliosphere, often associated with solar flares. The real time detection and analysis of these phenomena is of fundamental importance, since in their evolution they can collide with the Earth magnetosphere, causing geomagnetic storms and possibly disrupting human technologies, such as satellites or electrical power grids. Starting from sequences of C2 and C3 images acquired by the Large Angle and Spectrometric Coronagraph (LASCO), we want to detect the formation of CMEs and follow their evolution by the well-known Level-Set method, first introduced by Osher and Sethian [4] and successfully applied to many propagation problems over the years. Astronomical images are usually characterized by low resolution and significant presence of noise, due to the inherent limitations of the instrument sensors. Moreover, the high-intensity foreground luminosity emanating from the Sun makes the visualization of the CME a non trivial task. We propose to solve the problem by first filtering the input images to reduce the noise, then by applying some rescaling techniques to highlight the CME area [5]. Finally, we use a Semi-Lagrangian scheme [1, 3] for the Chan-Vese model [2] to segment the resulting images. The implemented method is efficient, stable and improves the state-of-the-art methods in terms of accuracy and reliability of the results. REFERENCES [1] Carlini E., Falcone M., and Ferretti R. Numerical Techniques for Level Set Models: an Image Segmentation Perspective. Level Set Methods in Medical Imaging Segmentation, Taylor & Francis (2019). [2] Chan T.F., and Vese L.A., Active contours without edges. IEEE Transactions on Image Processing (2001). [3] Falcone M., and Ferretti R. Semi-Lagrangian Approximation schemes for linear and Hamilton-Jacobi equations, SIAM (2014). [4] Osher S. and Sethian J. A. Fronts propagating with curvature-dependent speed: Algorithms based on Hamilton-Jacobi formulations. Journal of Computational Physics (1988), 79(1):12–49. [5] Tozza S., and Falcone M. On the Segmentation of Astronomical Images via Level-Set Methods. In: Donatelli, M., Serra-Capizzano, S. (eds) Computational Methods for Inverse Problems in Imaging. Springer INdAM Series, vol 36. Springer, Cham.