Digital borehole camera technology can accurately acquire information regarding the structural surface characteristics of rock within a borehole. To address the shortcomings such as labor-intensity, subjectivity, and computational intensity associated with existing digital borehole image analysis, this paper introduces a new analysis scheme to automate the recognition of borehole interior images captured by digital borehole camera technology. The proposed scheme begins by uniformly illuminating images using a two-dimensional gamma function light-adaptive correction algorithm. Next, edge features are extracted using a pre-trained DexiNed network. To tackle edge point noise and extract the region of interest, the Epremoval method is employed. Finally, the method performs polynomial fitting on the characterization data in the image utilizing the Taylor expansion of the sine curve. The parameters of the rock structure surface are obtained by calculation, spatial transformation and mathematical transformation of the obtained curves. The algorithm is applied to the digital borehole image of a tunnel project as an illustrative example. The obtained results are compared with the results of manual assisted interpretation, revealing superior recognition capabilities of the proposed method.