Ocean surface sunglint reflection serves as both unique information source and persistent data correction challenge in marine optical remote sensing. This study elucidates the multi-scale image characteristics and differentiation of sunglint reflections in remote sensing imagery. In coarse-resolution images, sunglint reflections exhibit statistical distribution characteristics. But high-resolution imagery reveals distinct discrete features, statistical features, and Fresnel reflection discrepancies. Current sunglint correction methods for marine optical imagery are systematically summarized and applied to multi-source remote sensing datasets. The application characteristics of different correction approaches are analyzed to discuss the effectiveness of sunglint reflection variations in detecting ocean surface targets, marine environmental dynamics phenomena, and environmental monitoring. Finally, prospects for future advancements in ocean sunglint remote sensing are outlined, providing a foundation for subsequent research endeavors.