Abstract: The Shenhu area in the South China Sea is one of the main areas rich in natural gas hydrate resources. Two on-site experimental explorations in 2017 and 2022 have confirmed the utilization prospects of hydrate resources. At present, precise evaluation on hydrate-containing reservoirs in the region yet needs further improvement as the microscopic occurrence of hydrates in sediment pore spaces is a key factor. This study addresses the issue of insufficient resolution in CT image representation of micro-occurrence forms of hydrates. A super-resolution reconstruction algorithm based on self-supervised deep-learning was established, in which a 2-fold and 4-fold increase in spatial resolution of CT scanning images were achieved. On this basis, the evolution of pore structure and microscopic occurrence characteristics of hydrates in the Shenhu area of the South China Sea were characterized. Due to the presence of a large number of foraminiferal shells in the sediments of the South China Sea, hydrates occupy mainly the internal space of the foraminiferal shells and block the connecting throats among pores, which significantly reduced the gas and water permeability of sediments. However, hydrates do not fully occupy the entire pore space, and there will still be a small amount of gas and water residue. Gas distributed mainly inside the hydrate particles, while water distributed mainly on the surface of hydrate particles. The above experimental results offered a guidance to the interpretation of on-site exploration data such as earthquakes and logging.