Systems and methods for processing geophysical data

Patent metadata

Field	Value
Publication number	US-9916539-B2
Application number	US-201314408957-A
Country	US
Kind code	B2
Filing date	Jun 18, 2013
Priority date	Jun 18, 2012
Publication date	Mar 13, 2018
Grant date	Mar 13, 2018

Abstract

Described herein is a computer implemented method for generating a probabilistic model usable to identify instances of a target feature in geophysical data sets stored on a memory device. The computer implemented method using a computer processing unit to generate a probabilistic model from a training library for use in identifying instances of the target feature in the geophysical data sets, applying, using the computer processing unit, the probabilistic model to one or more of the geophysical data sets to generate a plurality of results, processing, using the computer processing unit, the set of results according to an acceptability criteria in order to identify a plurality of candidate results, receiving a selection of one or more of the candidate results and for the or each selected candidate result displaying on a display the result and its associated geophysical data set to assist a user in making an assessment as to whether or not the probabilistic model is an acceptable model for the processing of the geophysical data sets, receiving from a user an assessment as to whether or not the probabilistic model is an acceptable model; and if the assessment received indicates the probabilistic model is an acceptable model for processing the geophysical data, outputting the probabilistic model and/or the training library.

First claim

The invention claimed is:

1. A computer implemented method for generating a probabilistic model usable to identify instances of a target feature in geophysical data sets stored on a memory device, the computer implemented method including:
(a) using a computer processing unit to generate a probabilistic model from a training library, the model for use in identifying instances of the target feature in the geophysical data sets, the training library including one or more target examples, each target example including a signature being indicative of the target feature, and one or more non-target examples, each non-target example including a signature being indicative of a non-target feature;
(b) applying, using the computer processing unit, the probabilistic model to one or more of the geophysical data sets to generate a plurality of results, each result associated with a processed geophysical data set and indicating a level of certainty as to whether that geophysical data set includes the target feature;
(c) processing, using the computer processing unit, the plurality of results according to an acceptability criteria in order to identify a plurality of candidate results, the candidate results being results associated with data sets having potential significance to the performance of the probabilistic model;
(d) receiving a selection of one or more of the candidate results and for each selected candidate result displaying on a display the result and its associated geophysical data set to assist a user in making an assessment as to whether or not the probabilistic model is an acceptable model for the processing of the geophysical data sets;
(e) receiving from a user an assessment as to whether or not the probabilistic model is an acceptable model; and
(f) if the assessment received indicates the probabilistic model is an acceptable model for processing the geophysical data, outputting the probabilistic model and/or the training library; and
wherein if the assessment received at step (e) indicates the probabilistic model is not an acceptable model for the processing of the geophysical data, the method further includes:
(g) receiving a selection of at least one example to be added to the training library, each example including a signature of either a target or non-target feature and being included in a data set associated with a candidate result, and modifying the training library by adding the at least one example; and/or
presenting the training library examples to the user, receiving a selection of one or more examples for removal from the training library, and modifying the training library by removing the example or examples selected for removal; and
(h) repeating steps (a) to (f) in respect of the modified training library.

2. A computer implemented method according to claim 1 , wherein using the computer processing unit to generate the probabilistic model from the training library includes applying Gaussian processes to the training library.

3. A computer implemented method according to claim 2 , wherein the acceptability criteria is based on a standard deviation of the plurality of results.

4. A computer implemented method according to claim 1 , wherein the acceptability criteria is based on the occurrence of results which cross a predetermined threshold when the results are considered in conjunction with the level of certainty.

5. A computer implemented method according to claim 4 , wherein the predetermined threshold is 0.5, and wherein consideration of a result in conjunction with the level of certainty includes:
if the result is greater than 0.5, subtracting the standard deviation to the result to see whether the resulting value is less than 0.5; or
if the output is less than 0.5, adding the standard deviation to the result to see whether the resulting value is greater than 0.5.

6. A computer implemented method according to claim 1 , wherein prior to adding the at least one example to the training library in step (g) the method includes, for each example selected to be added to the training library:
displaying a comparison of the selected example with one or more examples included in the training library to allow the user to make a compatibility assessment as to whether the selected example is compatible with the examples included the training library;
receiving an assessment as to whether the selected example is compatible with the examples included in the training library; and only modifying the training library by adding the selected example if the selected example is assessed as being compatible with the examples in the training library.

7. A computer implemented method according to claim 1 , wherein the training library is an initial training library and the method further includes: generating the initial training library by:
receiving a user selection of at least one target example from the geophysical data sets;
receiving a user selection of at least one non-target example from the geophysical data sets; and
adding the user selected target and non-target examples to the initial training library.

8. A computer implemented method according to claim 1 , wherein each geophysical data set is a natural gamma data log including natural gamma measurements taken from a drill hole.

9. A computer implemented method according to claim 8 , wherein the target feature to be identified in each natural gamma log is the existence of one or more marker shale bands.

10. A computer implemented method for identifying instances of a target feature in geophysical data sets stored on a memory device, the geophysical data sets having associated surface coordinates, the computer implemented method including:
implementing a computer implemented method according to claim 1 in order to generate a probabilistic model;
applying the probabilistic model to the geophysical data sets to generate a classification result for each data set, each classification result indicating whether the associated data set includes the target feature; and
outputting at least those classification results which indicate that the associated data set includes the target feature, together with information enabling the surface coordinates of the geophysical data set associated with the result to be determined.

11. A computer implemented method according to claim 10 , wherein a classification result is deemed to indicate that the associated data set includes the target feature if the result is above a classification threshold.

12. A computer implemented method according to claim 10 , wherein a classification result is deemed to indicate that the associated data set includes the target feature if the result is above a classification threshold after a standard deviation of the classification results is subtracted from the classification threshold.

13. A computer implemented method according to claim 11 , wherein the classification threshold is 0.5.

14. A computer implemented method for identifying the location of ore in a mining environment, said method including:
acquiring a plurality of geophysical data sets from the mining environment, each geophysical data set being associated with a drill hole having known surface coordinates;
implementing a computer implemented method according to claim 10 using said geophysical data sets to identify instances of a target feature occurring in the geophysical data sets;
using the classification results output from the computer implemented method according to claim 10 to predict a depth at which ore is likely to be found; and
outputting depth data representing said predicted depth together with information enabling surfac

Assignees

• Univ Sydney

Inventors

• Silversides Katherine Lee
• Melkumyan Arman
• Wyman Derek
• Hatherly Peter James

Classifications

• G06N7/01 Primary

  Probabilistic graphical models, e.g. probabilistic networks · CPC title

• G01V5/06

  for detecting naturally radioactive minerals · CPC title

• G06N20/00 Primary

  Machine learning · CPC title

• G06N99/005

  Physics · mapped topic

• G06N7/005 Primary

  Physics · mapped topic