Natural fractures provide important reservoir space and migration channels for oil and gas in tight reservoirs. Moreover, they are key factors controlling the high yield of tight oil and gas. Accordingly, methods to identify and characterize fractures are essential; however, conventional well-logging data are not ideal for such purposes. To this end, our study proposed an efficient method for identifying and characterizing fractures. First, core observations, core sample test analysis, numerical simulations, core calibration of borehole image logs, and borehole image log calibration of conventional logs were performed to identify sensitive log curves and log response characteristics of fractures. Second, we analyzed the response characteristics and differences in the log and properties of tight sandstone with and without fractures. Third, logging eigenvalue (EvF) and rock physical eigenvalue (MvF) models were constructed to determine different causes of fractures in tight sandstone. Finally, the two models were applied to identify and characterize fractures in the tight sandstone reservoirs in the West Sichuan Xinchang gas field, China. It was found that the effect of using the logging eigenvalue and rock physical eigenvalue models to identify fractures was similar to that observed using an image log. Overall, different fractures were accurately identified and characterized, indicating that the proposed method efficiently identifies and characterizes fractures in tight sandstone, ultimately advancing the research of fine reservoir evaluation and fracture theory.