Abstract:This study proposed a Beamlet-based method of extraction from core fissures images. Firstly, the collected core fissures images were transferred to the binary images using Otsu's thresholding segmentation algorithm, then the Beamlet transform was exploited to extract the linear features of fissures from the binary images with different scales and thresholds, and the relevant analytical data of fissures were added up according to pixel values. The results showed that the proposed method could extract clear images of core fissures in the cases of complex core backgrounds and noises, achieving satisfactory performance.