Abstract: Non-destructive testing (NDT) technology is a type of testing technology that evaluates and measures material properties without causing damage to the material or its structural performance. Despite the advancements in NDT technology, the detection of complex internal damages at multiple levels remains a significant challenge in the field. The objective of this study is to review and analyze the development, classification and challenges of NDT technology by combining domestic and international research results. On this basis, the study introduces a set of research results on a comprehensive and precise inspection system developed by the author's research team, which covers macro to mesoscopic levels and internal structures. For macroscopic surface identification and quantitative mesoscopic identification of apparent defects, the author’s team has developed visual detection technology. This technology involves rapid panoramic image stitching of spatial reference points and sub-pixel-level disease segmentation, as well as centimeter-level positioning methods for apparent defects. Additionally, the team proposed an artificial intelligence algorithm capable of simultaneously identifying multi-size micro-meso cracks ranging from 0.05 to 0.2 mm, and authenticating panoramic images. Regarding structural internal damage identification, the author’s team pioneered a new principle of intelligent variable-frequency acoustic hammering scanning, enabling precise identification of various types of damage. The team also established the theoretical method of damage detection and assessment using adaptive excitation distribution of mobile acoustic hammering, artificial intelligence algorithms for acoustic wave and acoustic image features, and key technologies for intelligent equipment. Experimental verification has shown that the maximum depth of crack detection is up to 40 mm with a width of 0.05 mm, the maximum depth of delamination detection is up to 400 mm and the minimum recognition range is 50 mm.