Bio-imaging plays an essential role in life sciences and medicine. Recently, structured illumination has emerged as a promising super-resolution technique. However, in the case of thick scattering media, the benefits of structured illumination can be significantly reduced compared to conventional illumination methods, resulting in significant degradation of imaging quality. Therefore, the question of dealing with the challenges posed by structured illumination in scattering media has become a topic of considerable focus. In this thesis, an experimental setup aimed at achieving super-resolution imaging using structured illumination is developed. The factors influencing the imaging process are systematically analyzed. At the same time, machine learning methods for recovering structured illumination patterns propagated through scattering media are explored. Overall, this research tackles the challenge of improving super-resolution imaging through scattering medium thereby contributing to the development of biological and medical imaging technologies.