Real-time pattern recognition and automatic interpretation of acoustic reflection images

1. A method for performing well logging in a borehole intersecting an earth formation, the method comprising:

obtaining an acoustic reflection image by:

generating waves that radiate away from the borehole into the formation, and

detecting reflections of the waves from at least one structural interface from outside the borehole traveling from the formation to the borehole,

the acoustic reflection image including the at least one reflecting structural interface of the formation;

automatically generating a compressed representation of the acoustic reflection image by:

identifying a set of features in the acoustic reflection image substantially fitting a pattern, wherein the set of features corresponds to a portion of the at least one reflecting structural interface of the formation; and

using a representation of the pattern as the compressed representation of the acoustic reflection image; and

transmitting the compressed representation uphole.

2. The method of claim 1 wherein the features are amplitude peaks in the acoustic reflection image, and the pattern is a line segment therein that is obtained from the amplitude peaks.

3. The method of claim 2 wherein identifying the set of features comprises generating a binary image of the amplitude peaks.

4. The method of claim 2 wherein the pattern is encoded by at least one of: i) a Hough transformation; ii) a modification of the Hough transformation.

5. The method of claim 2 wherein identifying the set of features in the acoustic reflection image substantially fitting the pattern comprises determining an envelope of the acoustic reflection image in a determined strike direction and finding dominant peaks in the envelope.

6. The method of claim 5 where the dominant peaks are found by a peak detection algorithm using automated global thresholding.

7. The method of claim 5 wherein the acoustic reflection image is obtained from an original acoustic reflection image via the application of a 2-D spatial wavenumber filter automatically generated as a polar segment within the 2-D spatial wavenumber amplitude spectrum of at least a portion of the original acoustic reflection image, and wherein an azimuthal orientation of the polar segment defines the determined strike.

8. The method of claim 1 , wherein the acoustic reflection image is obtained from an original acoustic reflection image via the application of a 2-D spatial wavenumber filter.

9. The method of claim 1 wherein the representation of the pattern results from conventional compression of an initial representation of the pattern.

10. The method of claim 1 comprising receiving the compressed representation at the surface in near real-time with respect to initiation of the transmitting.

11. The method of claim 1 comprising obtaining the acoustic reflection image with a low-frequency acoustic well logging tool.

12. A method for performing well logging in a borehole intersecting an earth formation, the method comprising:

obtaining an acoustic reflection image by:

generating waves that radiate away from the borehole into the formation, and

detecting reflections of the waves from at least one structural interface from outside the borehole traveling from the formation to the borehole,

the acoustic reflection image including the at least one reflecting structural interface of the formation;

automatically generating a compressed representation of an acoustic reflection image in the borehole by:

identifying a set of features in the acoustic reflection image by characterizing portions of the image meeting at least one predefined characteristic as a first type and characterizing all other portions of the image failing to meet the at least one predefined characteristic as a second type, wherein the set of features corresponds to a portion of the at least one reflecting structural interface; and

generating a binary image by representing the first type with first regions of the image and representing the second type with second regions of the image; and

using a representation of the binary image as the compressed representation of the acoustic reflection image;

transmitting the compressed representation uphole.

13. The method of claim 12 wherein the at least one predefined characteristic comprises an amplitude peak.

14. A well logging system for logging in a borehole intersecting an earth formation, the apparatus comprising:

an acoustic logging tool configured to be conveyed into a borehole;

a transmitter on the tool configured for generating waves that radiate away from the borehole into the earth formation;

at least one receiver on the tool configured for detecting reflections of the waves from at least one structural interface from outside the borehole traveling from the earth formation back to the borehole; and

at least one processor on the tool configured to generate an acoustic reflection image from the signal, the acoustic reflection image including the at least one reflecting structural interface of the formation, and automatically generate a compressed representation of the acoustic reflection image by:

identifying a set of features in the acoustic reflection image substantially fitting a pattern, wherein the set of features corresponds to a portion of the at least one reflecting structural interface of the formation; and

using a representation of the pattern as the compressed representation of the acoustic reflection image; and

a telemetry unit configured to transmit the compressed representation uphole.

15. The well logging system of claim 14 wherein the features are amplitude peaks in the acoustic reflection image, and the pattern is a line segment therein that is obtained from the amplitude peaks.

16. The well logging system of claim 15 wherein the at least one processor is configured to identify the set of features by generating a binary image of the amplitude peaks.

17. The well logging system of claim 15 wherein the pattern is encoded by at least one of: i) a Hough transformation; ii) a modification of the Hough transformation.

18. The well logging system of claim 15 wherein the at least one processor is configured to identify the set of features in the acoustic reflection image substantially fitting the pattern by determining an envelope of the acoustic reflection image in a determined strike direction and finding dominant peaks in the envelope.

19. The method of claim 1 comprising obtaining the acoustic reflection image with at least one of: i) a monopole acoustic well logging tool, and ii) a dipole acoustic well logging tool; iii) a quadrupole acoustic well logging tool.

20. The method of claim 14 wherein the representation of the pattern results from conventional compression of an initial representation of the pattern.

21. A method for performing well logging in a borehole intersecting an earth formation, the method comprising:

obtaining an acoustic reflection image by:

generating waves that radiate away from the borehole into the formation, and

detecting reflections of the waves from at least one structural interface from outside the borehole traveling from the formation to the borehole,

the acoustic reflection image including the at least one reflecting structural interface of the formation;

automatically generating a compressed representation of the acoustic reflection image by:

identifying a set of features in the acoustic reflection image substantially fitting a line segment pattern, wherein the set of features corresponds to a portion of the at least one reflecting structural interface of the formation; and

using a representation of the line segment pattern as the compressed representation of the acoustic reflection image; and

transmitting the compressed representation uphole.