Skip to main content

Data Augmentation Module Per Data Type

Updated yesterday

Overview

Different Data Augmentation Modules are designed to work with specific types of geoscientific data. Understanding which Module to use, and what kind of input or output it expects, helps streamline preprocessing and improves performance in DORA's modelling workflows.

This article summarizes which Modules are best suited to geology, geophysics, and geochemistry, and how each handles inputs and outputs. Some Modules generate one raster per input (1 to 1), others process multiple inputs into a single output (many to 1), and some split a single input into multiple outputs (1 to many).

By Data Type

Geology and Structure

Module

Input Type

Input to Output Relationship

Notes

Distance Maps

Vector Data

1 per type → 1 output per type

One raster output per feature category, such as each fault

Structural Field Maps

Vector Data

1 per type → 1 output per type

Orientation-based inputs

Fault Disturbance Maps

Vector Data

1 input → 1 output

Calculated using faults and input parameters

Geological Domain Probability Maps

Vector Data

(Polygon)

Multiple domains → 1 output

All domains create a single raster, i.e., a lithology map.

Geophysics

Module

Input Type

Input to Output Relationship

Notes

Computer Vision Maps

Raster

1 inputs → multiple outputs

Deep learning (ResNet); single-layer input

Texture Filter Maps

Raster

1 input → 4 outputs

Outputs include rasters for Contrast, Correlation, Energy, and Entropy.

Lineament Maps

Raster

1 input → 2 outputs

Outputs include Line Density and Structural Complexity

Geochemistry

Module

Input Type

Input to Output Relationship

Notes

Rock Geochemical Maps

Point Data

1 input → multiple outputs; 1 per element

One raster per geochemical element

Multivariate Anomaly Maps

Raster

Multiple inputs → 1 output

Identified anomalies from the input maps.

Surficial Sediment Geochemical Maps

(Streams, Soil, Seds)

Point Data

1 input → multiple outputs per element

One raster per geochemical element

Other

Module

Input Type

Input to Output Relationship

Notes

Data Density Maps

Any spatial data (Vector or Raster)

Multiple inputs → 1 output

Counts the number of datasets contributing to each pixel

VRIFY Prospectivity Score (VPS)

Feature stack from all Modules

Multiple inputs → 1 output

Final predictive map based on selected features


Learn more


Still have questions?

Reach out to your dedicated DORA contact or email support@VRIFY.com for more information.

Did this answer your question?