Abstract: In order to increase the efficiency of rock information collection, this study focuses on tackling the issue of acquiring rock information via digital image techniques, integrating sample windows with image processing techniques. In the research, gaussian adaptive filtering completed image noise reduction, which can reduce the influence of noise on the image. Canny algorithm extracted based on boundaries, the images were then improved using distance thresholding, angle thresholding, and length thresholding. At the same time, according to angles grouped the fracture images, morphological algorithms obtained the number of fractures. Through practical use in a mine in Shengzhou, Zhejiang Province, the viability and usefulness of this digital image processing approach in geological investigations were confirmed. The study is of some significance for the effective capture of rock information because the proposed method of fracture extraction shows accuracy levels higher than 93%.