Abstract: This article systematically reviews the research progress and application status of machine learning in the field of metallographic analysis.It emphatically analyzes the technological breakthroughs of machine learning in key links,including grain size rating,microstructure identification,non-metallic inclusion analysis,case depth detection and intergranular corrosion assessment.Meanwhile,the existing issues are pointed out,such as uneven data quality,insufficient algorithm applicability and poor model interpretability.Combined with the development of artificial intelligence(AI) technology,future development directions are prospected,including the construction of standardized databases,the development of multimodal learning frameworks and the improvement of model interpretability.This study aims to promote the evolution of metallographic analysis towards intelligence and standardization,and provide stronger technical support for material research and quality control.