Feature detection in seismic volumes

A computing system, comprising:

• one or more processors;
• and
• one or more computer readable media storing instructions that, when executed by at least one of the one or more processors, are configured to cause the computing system to perform operations, comprising:
• performing a windowed Radon transform analysis of a domain, wherein performing the windowed Radon transform analysis of the domain comprises:
• defining a mask plane comprising a first dimension of a first number of voxels and a second dimension of a second number of voxels;
• and
• selecting a plurality of first angles for orientating the mask plane in the domain with respect to a first axis, wherein for each one of the plurality of first angles selected:
• populating, using one or more processors, sum cubes associated with each one of a first plurality of subject voxels, wherein populating the sum cubes for each of the first plurality of subject voxels comprises:
• defining a first line associated with the one of the first plurality of subject voxels, the first line extending at the one of the plurality of first angles, the first line intersecting the first number of voxels including the one of the first plurality of subject voxels;
• and
• populating the one of the sum cubes associated with the one of the first plurality of subject voxels with an intermediate sum of values associated with each of the first number of voxels intersected by the first line, wherein for at least some of the first plurality of subject voxels, populating comprises using a running sum, wherein using the running sum comprises:
• determining the intermediate sum calculated for a previous one of the first plurality of subject voxels, wherein the previous one of the first plurality of subject voxels is displaced from a current one of the first plurality of subject voxels by a third number of voxels and the current one of the first plurality of subject voxels is intersected by the first line associated with the previous one of the first plurality of subject voxels;
• and
• subtracting one or more values of one or more voxels intersected by the first line associated with the previous one of the first plurality of voxels but not intersected by the first line associated with the current one of the first plurality of voxels;
• and
• after populating the sum cubes, selecting a plurality of second angles, wherein for each one of the plurality of second angles selected, calculating, using the one or more processors, a planar sum for each one of a second plurality of subject voxels, wherein calculating the planar sum for each one of the second plurality of subject voxels comprises:
• defining a second line extending at the one of the plurality of second angles, intersecting the second number of voxels including the one of the second plurality of subject voxels;
• and
• calculating a planar sum associated with the one of the second plurality of subject voxels by calculating a sum of values corresponding to the intermediate sums associated with each of the second number of voxels intersected by the second line;
• and
• detecting a feature in the domain using the planar sum.

A non-transitory computer readable medium storing instructions thereon that, when executed by a processor, are configured to cause a computing system to perform operations, comprising:

• performing a windowed Radon transform analysis of a domain, wherein performing the windowed Radon transform analysis of the domain comprises:
• calculating intermediate sums for a plurality of first subject voxels, wherein for each one of the first plurality of subject voxels, the one of the intermediate sums associated therewith corresponds to a sum of values associated with voxels that neighbor the one of the first plurality of subject voxels, wherein calculating the first sums comprises:
• storing the one of the first sums calculated for a first one of the first subject voxels as a running sum;
• subtracting one or more first values associated with one or more voxels from the running sum;
• and
• adding one or more second values associated with one or more voxels from the running sum;
• and
• calculating a planar sum for a second plurality of subject voxels, wherein for each one of the second plurality of subject voxels, the one of the planar sums associated therewith corresponds to a sum of intermediate sums associated with voxels that neighbor the one of the plurality of second subject voxels;
• and
• detecting a feature in the domain using the planar sum.