Mangrove forests are vital and inseparable components of coastal ecosystems, primarily thriving in tropical regions. This study conducts a comparative assessment of a dedicated mangrove product, land cover products, spectral indices, and the Support Vector Machine (SVM) algorithm, with a particular focus on the identification of the mangrove class. For this purpose, 10-m spatial resolution products including High-resolution Global Mangrove Forests (HGMF), ESA WorldCover, ESRI Landcover, the Mangrove Vegetation Index (MVI), and the SVM classification algorithm were applied to two study areas: northwestern Australia and southern Iran (Sirik region). The results indicate that the HGMF and ESRI products in the Sirik region yielded lower Kappa coefficients (0.59 and 0.32, respectively) compared to their counterparts in Australia (0.93 and 0.61, respectively). In contrast, the ESA product, along with the remote sensing derivatives SVM and MVI, achieved higher Kappa values in Sirik (0.75, 0.92, and 0.82, respectively). Meanwhile, in northwestern Australia, the same remote sensing derivatives reached Kappa values of 0.86, 0.98, and 0.73, demonstrating significantly higher accuracy. In general, the findings suggest that these products exhibit greater sensitivity to extensive and close mangrove formations. The high density of such mangroves in the Australian study area contributed to higher classification accuracy. In contrast, the presence of both dense and sparse (open) mangroves in the Iranian region reduced classification performance, as sparse mangroves were less accurately detected by the products. On the other hand, machine learning algorithms effectively identified both close and open mangroves, likely due to the supervised nature of their classification. However, the MVI index caused pixel mixing in the study areas, resulting in the misclassification of some non-mangrove regions as mangrove areas.