Rapid post-eruption mapping is needed to provide initial information about the impact of an eruption. Rapid mapping of eruption impacts using optical data is often tricky due to cloud cover, so combining optical data with SAR data is essential. This paper presents an object-based machine-learning procedure that utilizes Synthetic Aperture Radar (SAR) from Sentinel-1 and optical data from Sentinel-2 data to rapidly map recent volcanic deposits in the 2022 Semeru Volcano eruption. This study aims to test the algorithm’s performance in identifying recent volcanic deposits during the eruption. NDVI (Normalized Difference Vegetation Index) and NHI (Normalized Hotspot Index) spectral index changes combined with Sentinel-1 SAR information in the form of interferometric coherence, VV and VH beta nought values, and polarimetric decomposition SAR features are used as model inputs. The result indicated that combined SAR and optical data use in an object-based Random forest algorithm provides good recent volcanic deposit detection. The algorithm’s performance detecting the recent volcanic deposit in the Semeru Volcano case shows a good Dice Coefficient of 0.73 and an F1-score of 0.83. This straightforward algorithm will improve the rapid mapping process during and after volcanic eruptions.