Context. Movies of the solar atmosphere reveal motion and variations in brightness. In particular, sequences of coronal images exhibit the plane-of-the-sky component of the velocity combined with other variations in the signal. The present work analyses solar extremeultraviolet images, as recorded by the Extreme ultraviolet Imaging Telescope (EIT) on board the Solar and Heliospheric Observatory (SoHO) and by the Transition Region and Coronal Explorer (TRACE). Aims. Our aim is to simultaneously estimate the apparent motion vector and the variation in brightness from two successive images.Methods. We present a multiscale optical-flow algorithm derived from a local gradient-based technique that estimates the deformation parameters. Our algorithm is symmetric in the sense that it computes the exact same estimation if the two images are swapped. This also regularises the optical flow when two local image patterns do not match, e.g. in case of temporal sub-sampling. Independently our algorithm regularises the optical flow against aperture effects occurring typically along coronal loops. Results. We demonstrate a new differential rotation measurement technique and the identification of coronal events as regions exhibiting a significant brightness variation or an outstanding velocity field. Space weather services have motivated this study. The range of potential interests includes, but also extends beyond, on-disc signatures of flares and coronal mass ejection (CME). It encompasses, for example, studies of bright points and filaments, coronal seismology, and EIT wave investigations.