Illustration of Euclid’s capabilities to measure galaxy morphologies. Top panels: example of a simulated galaxy observed with VIS as compared to HST and Subaru/HSC. The horizontal black line indicates a 1″ length. Middle panels: comparison of the bias (left column), dispersion (middle column) and outlier fraction (right column) of the effective radii (top row), axis ratio (middle row) and Sérsic index (bottom row) for the best-fit Sérsic profiles obtained with different state-of-the art surface brightness fitting codes applied to simulated Euclid galaxies as a function of I_E. Sérsic parameters can be obtained with errors smaller than ~10% down to a I_E = 24. Bottom panels: accuracy of deep learning based morphological classifications on simulated Euclid observations of galaxies trained on human based labels. The confusion matrices show the accuracy for identifying spiral arms (left) and clumpy galaxies (right). Figure adapted from Euclid Collaboration: Aussel et al. (2024) and Euclid Collaboration: Bretonnière et al. (2023).