Fractures of different scales are developed in the tight sandstone reservoirs of the Upper Triassic Yanchang Formation in the Ordos Basin, which are complexly distributed and difficult to predict. In this paper, taking the tight sandstone reservoirs of the Upper Triassic Chang 6 Oil Group in the Wuqi area of Northern Shaanxi Province as an example, based on the data of conventional logging, imaging logging, core observation, and cast thin sections, an improved variable scale fractal method combined with second derivative of lg(R/S) and grey relational analysis was used to calculate the fracture development coefficient Q established on five types of conventional logs. By integrating this with imaging logging data, we determined the lower limit value of Q indicative of fracture development and quantitatively evaluated the fracture development in the Chang 6 tight sandstone reservoirs of the study area. The results show that the reservoirs in the Chang 6 Oil Group of the Yanchang Formation in the Wuqi area mainly develop vertical fractures and high-angle fractures, and the lower limit of the value of the fracture development coefficient Q for this set of reservoirs is 0.8, and the locations of fracture development. assessed by this method show a high degree of agreement with the fractures interpreted from the imaging logs. Additionally, the fracture development coefficient Q evaluated by this method is positively correlated with fracture line density, with a high degree of correlation. The improved variable scale fractal method, compared to the conventional variable scale fractal method, has higher evaluation accuracy in identifying fractures for the tight sandstone reservoirs. It is widely recognized that the improved variable scale fractal method for quantitative evaluation of fractures in tight sandstone reservoirs has better reliability compared to conventional methods. So, this study can provide a new approach for predicting natural fractures in tight sandstone oil reservoirs.