Coral reef pore recognition and feature extraction based on borehole image

The coral reef pores are the main reservoir space of oil and gas resources. Through determining the structural form and distribution characteristics of pores, it can be helpful for scientific and reasonable stratigraphic division and reservoir evaluation, and provide basic data for the efficient development of using coral reefs as oil and gas reservoir sites. The core difficulty of interpretation and evaluation of coral reef pore structure and distribution characteristics is the pore recognition and feature extraction. In view of problems of low accuracy and efficiency in the recognition and feature extraction of coral reef pore at present, this paper proposes a method of coral reef pore recognition and feature extraction based on borehole image data. First of all, based on the borehole image and combined with the pore characteristics in borehole image, this paper proposes the coral reef pore recognition method. By using the feature of color gradient difference between the pore and the non pore in the borehole image, it realizes the segmentation of coral reef pore with low noise and weak interference. On the basis of obvious coral reef pores, the contour tracking method and image filling system of coral reef pore are established. Then, on the basis of traditional eight connected region marking algorithm, a pore region feature extraction method is proposed. The equivalent relationship between the temporary mark and the final mark is established, and an evaluation criterion that can comprehensively reflect the pore region feature parameters in the borehole image is formed. Finally, a systematic analysis and a comparison of results are carried out based on the actual coral reef borehole images. The results are consistent with the real values, which verifies the correctness of method.

Automated summary

The study proposed a method of coral reef pore recognition and feature extraction based on borehole image data, achieving pore segmentation through color gradient difference feature and establishing pore contour tracking and filling system.