What Are the Best Practices for Core Photography?

Once just for record-keeping, drill core photography is now valuable for analysis with new analytics. Datarock’s guidelines ensure images are high-quality, consistent, and easy to capture.

Timing of photography

Overview

The timing of core photography, relative to other steps in the core processing workflow, will affect the downstream analytics.  

• If geotechnical or textural analysis is required, the core should be imaged as early as possible (even at the drill rig).  Modification to the core, through transport or human annotation, should be minimised. 
• If sampling and geological observations need recording, the core should be photographed when the markup is complete. The impact of the additional annotations on the core can be minimised (see Section 3); however, there will typically be detrimental effects on future analytics. 
• We do not recommend photographing the core after it has been cut unless there is an earlier photo. 

Taking multiple photographs of a core tray

It is now becoming more common for companies to take multiple photographs of core trays as they are processed. A typical workflow might include

1. At the drill rig: An early photograph immediately after drilling; 
2. After metre marks have been drawn and core loss has been measured: This is the preferred stage as Datarock processing can assist logging;
3. After Logging: Photography after logging and markup; and
4. Final: A final photograph after cutting and sampling.

This workflow provides a comprehensive record of the core processing workflow, the best-case scenario for analytics, and an audit trail of any changes or damage to the core in transit. 

 

---

 

Preparation of core trays for photography

To maximise the potential benefit of automated extraction of information from core photography, we recommend several core tray preparation steps.

Writing on the core tray

Certain information should always be written on the core trays, preferably in a consistent location and format:

• To and from depths, preferably in the top left for the ‘to’ value, and bottom right for the ‘from’ depth - this aids in the quick adding of metadata later, either manually or through optical character recognition (OCR), 
• Drill hole name - consistent with no underscores, 
• Core tray number - 001 format
• The word “START” written on the top side of the tray in a consistent location - this will aid in programmatically identifying photos taken upside down or rotated later.

Example core tray markings including START and END depths, tray number and hole ID in the desired format.

Writing on core

Writing on the core is best completed with permanent markers, wax markers or chinograph pencils as they are the most durable and do not wash off when undertaking wet photography. 

Writing can generally be broken down into three categories:

• Metre marks should be written in white or black (depending on core colour), with large, clear writing. The sample line should always be to the left of the numbers and extend from the top to the bottom of the core. This will help minimise the optical character recognition confusing it with a ‘1’.
• Sample marks  should be written in red and will be differentiated from metre marks based on colour. The numbers will be written in the same format as metre marks. 
• Lithological boundaries should be marked in yellow. Lithological metadata should be written on the left-hand side of the boundary, and depth information on the right. 

In general, additional writing on the core should be minimised as it obscures the rock texture. If required, use distinct colours, which will not be confused with the three categories listed above, and consider taking an additional photo after metre marks have been drawn on and prior to sampling and lithology markings.

 

 

Orientation lines

Orientation lines should be aligned with the bottom of the tray channel but still visible to the camera. Ideally, this line will be marked with a permanent marker, wax marker or chinograph. Importantly this marking should remain consistent to allow automated processes to work accurately. The base of the channel is preferred as it allows the orientation line to be imaged but doesn’t obscure the rock texture.

Example: The core is aligned along the bottom of the core tray row, with the ticks extending at most 25% of the exposed rock (10-20% is ideal). 

Joints

Joints should be aligned but opened slightly (~0.5 cm) to allow the edges of the fractures to be clearly imaged by the camera.

Example of a well presented, natural joint. 

Drillers’ breaks

Drillers’ breaks should be marked on the core with a visible cross (‘X’) within 5 cm of a mechanical break, including at the end of a row if it is a mechanical break.

The colour can change depending on the colour of the rock (red should be the default), but the crosses must be clear and visible to the camera. When multiple mechanical breaks occur close together, pairs should be marked closer together to aid the algorithm in attaching them to the correct break (see below).

 

Example: Crosses are located on both sides of the break. If this is done correctly when the core is in its original location, movement and rotation of the core during transport can be tracked. 

Use of colour bar and scale card in photograph

Variations in lighting can be corrected using a consistent colour bar. We recommend the “Calibrite ColorChecker Classic”, shown below and available from a number of suppliers. It is suitable to place within the frame, or can be cut and placed in a long strip along the edge of the core tray.

Scale cards should also be used when measurements from images are required. 

In addition, we also recommend including a header sheet such as the one shown below to provide a standardized format and a prompt for personnel to include tray data.

---

 

Photo Acquisition

While there are a number of proprietary core imaging systems coming to market that take very high quality photos, there are some relatively simple and inexpensive steps that can be taken to significantly improve the quality and consistency of core photos.

The following sections provide suggestions to the various aspects of core photo acquisition. 

Lighting source

The type of light used to illuminate the core tray will play a significant role in the quality of the image. This is applicable to both general core shed lighting and controlled source lighting within a photography system.

A key metric to understand when choosing a lighting type is the colour rendering index (CRI). CRI is an attribute of the light source, and is a quantitative measure of the ability of a light source to reveal the colours of various objects faithfully in comparison with an ideal or natural light source. Light sources with the highest possible CRI should be used (https://en.wikipedia.org/wiki/Color_rendering_index).

 

 

Our recommendation would be to use lighting with a CRI measurement of at least 80. These are generally inexpensive, modular and remain relatively low temperature compared to incandescent lighting sources. Your electrical supplier will be able to readily source high CRI lights.

Light diffusers

Intense point source lights can cause reflections on the drill core, especially if the core is wet. Light sources (preferably high CRI, see above) should be diffused using an opaque shroud and ideally not be located directly above the core tray. It is best to avoid positioning light sources behind the camera, keeping them off at an angle/to the side of the centre of the tray as this will reduce the reflections commonly encountered in wet core photography (see example below).

 

Example: Reflections on core can mask important geological features, and can also be confused with core markups such as ori and cut lines.

 

There are cost effective camera solutions on the market that utilise core shed lighting and use a diffuser to minimise glare. An example is shown below: 

Casper System (external link)

Removing non-controlled light sources

The next step for lighting control is to remove non-controlled light sources to ensure consistency. We recommend an indoor environment and blockout shroud  to remove all background lighting. This way, outside light (often fluorescent) from the coreshed or sun can be completely removed and replaced with a controlled light source inside the system (discussed above). It is important that these products do not induce reflections on the core. Here are some relatively simple options:

Camera Position

The camera should be positioned directly over the core tray to ensure all of the core is included in the image. 

If it is taken from a high aspect ratio, or if multiple trays are captured in one image, the tray can obscure the sides of the core.

Polarised lens

When the core is photographed wet, it can be challenging to avoid reflections. One potential way to remove these is by using a polarising lens. The downside is that these lenses can reduce the overall amount of light reaching the sensor, meaning more powerful lighting sources are required.

Resolution

It is important to take photos at a high enough resolution that captures sufficient detail for current and future use cases. It is good practice to aim for around 100 micron resolution core photograph.

Modern, full-frame DSLR cameras generally have between 30 and 50 megapixels. Single-shot systems positioned approximately 1-1.5 metres above the core tray will produce images with approximately 100-150 micron resolution.

It is possible to position the camera closer to the core tray and take several photographs that are then stitched together, however this can be challenging and often leads to distortion and artefacts in the stitched image. We don’t recommend this approach as these issues often outweigh the benefits. 

Whatever camera is being used, it should always be set to capture its maximum image quality.

Focal length / Zoom

As the camera should be held in a fixed position with respect to the core tray, the focal length, or zoom, of the lens should generally also be fixed.

Aperture / Depth of Field

Sometimes adjustments in aperture are required when switching from full core to half core, these settings should be recorded and changed when required.

However the recommended settings below should be suitable for the common range of imagery.

Focus

To minimise blurry images, use auto focus or fix the focus for every image.

Recommended Camera Settings

Unfortunately there is no one set of camera settings that work best for all rock types. Very dark rocks can require different settings to very light rocks, and as with most things there are trade offs to be made with some settings. Get in touch if you need some advice on best practice camera set ups for your particular situation.

Fully manual settings will yield best results but will take some experimentation depending on the rocks, lighting conditions and also the exact camera and lens being used. 

The table below provides a good starting point.

 

Image Extent

The photography system should be set up to capture the entire core tray on all sides, with one tray per image. 

Do not cut out any of the tray as it often contains data such as depths, and can also present problems to the various machine learning models. 

In general, a higher quality image further away will be of more value than a close, low quality image, particularly with bad stitching.

 

Example of a good image extent.

 

Example of a poorly, manually overcropped image.

 

---

 

Photograph naming convention (image metadata)

To ensure images are uploaded to the platform and ingested correctly, it is essential that the metadata are recorded consistently according to the below standards. Critically,  From and To depths for each core tray should be recorded in the file names of each image.

File names are to be delimited by underscores ‘ _ ‘, containing information on hole ID, tray number(s), the From and To depth at the start and end of the tray and whether the hole is wet or dry. 

The minimal metadata recommended is:

HoleID_TrayNumber_FromDepth_ToDepth

Example: DDH1_001_5.00_10.65

Should photos have two core trays in one photo, a ‘+’ will be used to note this. The recorded depths should cover both trays (i.e. the start of the first tray and the end of the second tray). 

Example: DDH1_01+02_5.00_15.00

To identify whether a photo was taken wet or dry a _W or_D can be added. 

Example: DDH1_01+02_5.00_15.00_W

It is preferable that this metadata is collected during the photography process and entered into some sort of simple application. Most modern DSLR cameras have the ability to speak to 3rd party applications. Please speak to us if assistance is required creating a custom application to record the photography metadata.

 

---

 

Core loss

To accurately determine downhole depth, Datarock incorporates core loss. The Datarock Core accounts for this by allowing the user to upload a simple csv with core loss intervals. It is therefore important to ensure depth related core loss is accurately captured by incorporating into the logging and photography workflow.

If core loss is not recorded downhole, the platform will highlight areas where there is not enough rock to account for a given interval length, however it is best practice to capture this data. 

Example:  Accepted file naming and csv format.  Note that core loss data should only include depth related core loss, and not volume recovery.

 

---

 

Cheat sheet for optimising core imagery

Core tray preparation

1. Ensure core is neatly prepared and curated before the photograph
2. Take photos of whole core where possible (in general)

Core tray markup

1. Core should be marked with start (depth) and end (depth), Tray ID & Hole ID clearly in view of camera
2. Metre marks should be clearly marked in a distinctive way - avoid drawing marks that look like a 1 (this can limit the ability to use OCR)
3. Use colour stickers or other types of unique markers to identify blocks and metre marks if making use of automated image analysis methods
4. All markings including orientation should be displayed face up so it can be clearly captured in the image

Photo rig setup

1. Fixed camera and tray position
2. Take photos under a controlled and consistent light source
3. Use a booth to remove outside light sources
4. Use physical guides to place the core trays in the same position

Camera settings

1. Ensure correct camera settings
2. Have lens cleaning kit on hand, regularly check and clean lens 

Wetting the core

1. Use a low pressure mister to wet the core to avoid the formation of large droplets that can obscure the rock and create reflections
2. Take both dry and wet photos where possible

Metadata collection

1. Depth metadata is most important and should added to the file name of the photograph
2. The data and name of the user are also good
3. Use an automated system to enter this data (let us know if you need advice on this)

QAQC

1.  Have a set up that allows the user to easily examine the photos they are producing to check for defects

Typical core photography setup

---

 

Copyright

This document is the creation of Datarock Pty Ltd. © 2022. 

Disclaimer

The information contained in the submission is given in good faith and in the belief that it is not false or misleading, as at the date of this document.