Cracking in tunnel lining structures is a critical issue requiring effective prevention and control during tunnel maintenance. The development of intelligent tunnel management and maintenance urgently requires a quantitative evaluation method for tunnel lining cracking diseases that is practical and broadly applicable. This study introduces an automatic crack image recognition technology and establishes a quantitative evaluation index system and threshold determination method for both single crack diseases and multi-crack diseases in tunnel lining sections. This is achieved by integrating the product scale method with established specification requirements. For single crack disease diagnosis, the primary evaluation indexes are crack length and width, with depth, orientation, and development serving as auxiliary indexes. For multi-crack disease diagnosis, the main indexes remain crack length and width, while depth, orientation, development, and distribution density are included as supplementary indexes. Threshold determination incorporates the probability distribution characteristics of uniform distribution functions alongside the safety characteristics of tunnel structures under stress conditions. The evaluation scores have been systematically standardized to eliminate dimensional inconsistencies, and software has been developed for practical application. A case analysis shows the scientific validity of this method, providing a novel approach to intelligent identification of tunnel cracking diseases and automated maintenance decision-making.