The interaction of incoming barotropic semidiurnal tides from the western Pacific with the double ridges in the Luzon Strait generates large-amplitude baroclinic internal tides, which subsequently evolve into strongly nonlinear internal solitary waves (ISWs) in the northern South China Sea through nonlinear steepening. Synthetic Aperture Radar (SAR) and optical remote sensing can only capture the 2D distribution of sea surface roughness or reflectivity caused by internal solitary waves, which inevitably leads to errors when using satellite images to invert internal solitary waves. The Surface Water and Ocean Topography (SWOT) satellite, launched in 2022 and equipped with the Ka-band Radar Interferometer (KaRIn), marks a transformative advance in oceanographic remote sensing. However, improving the retrieval accuracy of ISWs from SWOT data critically depends on the development of robust dynamic decomposition algorithms, precise crest-line extraction methods, and accurate inversion of subsurface wave-current fields. In this study, we propose an automatic method for identifying and extracting ISW crest lines from SWOT observations, establish an inversion framework for ISW parameters and their associated velocity fields based on a strongly nonlinear model, and validate the inversion accuracy using in-situ measurements. On this basis, we analyze the characteristics of ISWs in the northern South China Sea using our proposed identification and retrieval schemes. The results provide a scientific foundation for leveraging SWOT's high-precision observational data to conduct comprehensive, high-resolution investigations into the spatiotemporal evolution and underlying dynamic mechanisms of ISWs in the South China Sea.