Recently, many label-free optical imaging methods based on the interaction between light and matter have been gradually developed and play an important role in more and more fields. Stimulated Raman microscopy, as an label-free vibrational imaging tool, has the advantage of high-resolution chemical specific imaging of the main biomolecules most relevant for histological diagnosis. In this article, we mainly focus on the technological development of stimulated Raman microscopy and its application in fast tissue imaging without staining or sectioning. After introducing optical principles and technical details, we demonstrated the application of stimulated Raman histology in various types of human diseases. Further combining deep learning algorithms, case studies on fresh surgical tissue, gastroscopy biopsy, and core needle biopsy were summarized to demonstrate the potential of artificial intelligence assisted stimulated Raman histology in various application.