Automatic Dip Picking From Azimuthal Borehole Images

Abstract

A method comprising: disposing an azimuthal borehole measurement tool into a borehole, obtaining an azimuthal borehole image with the azimuthal borehole measurement tool, running a pre-processing action on the azimuthal borehole image to obtain a smart gradient image, running a sinusoidal pattern search action on the smart gradient image to form a raw dip of the azimuthal borehole image, and running a dip post-processing action to form a final dip. Additionally, a system may include comprise: an azimuthal borehole measurement tool disposed in a borehole configured to obtain an azimuthal borehole image. The system may also comprise an information handling system configured to perform the method described above.

1 . A method comprising:
 disposing an azimuthal borehole measurement tool into a borehole;   obtaining an azimuthal borehole image with the azimuthal borehole measurement tool;   running a pre-processing action on the azimuthal borehole image to obtain a smart gradient image;   running a sinusoidal pattern search action on the smart gradient image to form a raw dip of the azimuthal borehole image; and   running a dip post-processing action to form a final dip.   
     
     
         2 . The method of  claim 1 , wherein the pre-processing action comprises data quality control performs a quality check on the azimuth borehole image. 
     
     
         3 . The method of  claim 2 , wherein the pre-processing action comprises quality-based resampling is performed on the azimuthal borehole image with data quality control to form a resampled azimuth borehole image. 
     
     
         4 . The method of  claim 3  further comprising taking a difference of data values between two adjacent data points along a depth direction or using a gradient filter. 
     
     
         5 . The method of  claim 1 , wherein the sinusoidal pattern search action comprises a non-uniform Hough transform to search for dips in the smart gradient image. 
     
     
         6 . The method of  claim 5 , wherein the non-uniform Hough transform is performed in a three-dimensional domain to search for sinusoids with different depths, amplitudes, and phases. 
     
     
         7 . The method of  claim 6 , wherein the Hough transforms implements median based voting, wherein median based voting aggregates the median value of values comprising gradients along a sinusoid and forms a 3D Hough volume. 
     
     
         8 . The method of  claim 7 , further comprising recording amplitude and phase of a maximum voting for each depth of the 3D Hough volume. 
     
     
         9 . The method of  claim 1 , wherein the dip post-processing action comprises determining one or more dip features, wherein one or more dip features comprises symmetry, separation, or dip continuity. 
     
     
         10 . The method of  claim 9 , further comprising computing a score for the raw dip based at least in part on the one or more dip features. 
     
     
         11 . A system comprising:
 an azimuthal borehole measurement tool disposed in a borehole configured to obtain an azimuthal borehole image; and   an information handling system configured to:
 run a pre-processing action on the azimuthal borehole image to obtain a smart gradient image; 
 run a sinusoidal pattern search action on the smart gradient image to form a raw dip of the azimuthal borehole image; and 
 run a dip post-processing action to form a final dip. 
   
     
     
         12 . The system of  claim 11 , wherein the pre-processing action comprises data quality control performs a quality check on the azimuth borehole image. 
     
     
         13 . The system of  claim 12 , wherein the pre-processing action comprises quality-based resampling is performed on the azimuthal borehole image with data quality control to form a resampled azimuth borehole image. 
     
     
         14 . The system of  claim 13 , wherein the information handling system is further configured to take a difference of data values between two adjacent data points along a depth direction or using a gradient filter. 
     
     
         15 . The system of  claim 11 , wherein the sinusoidal pattern search action comprises a non-uniform Hough transform to search for dips in the smart gradient image. 
     
     
         16 . The system of  claim 15 , wherein the non-uniform Hough transform is performed in a three-dimensional domain to search for sinusoids with different depths, amplitudes, and phases. 
     
     
         17 . The system of  claim 16 , wherein the Hough transforms implements median based voting, wherein median based voting aggregates the median value of values along a sinusoid and forms a 3D Hough volume. 
     
     
         18 . The system of  claim 17 , wherein the information handling system is further configured to record amplitude and phase of a maximum voting for each depth of the 3D Hough volume. 
     
     
         19 . The system of  claim 11 , wherein the dip post-processing action comprises determining one or more dip features, wherein one or more dip features comprises symmetry, separation, or dip continuity. 
     
     
         20 . The system of  claim 19 , wherein the information handling system is further configured to compute a score for the raw dip based at least in part on the one or more dip features.