Edited from the video's English captions.
0:00Introduction
Welcome to Neuralog. Today I'd like to introduce you to AILFE, an exciting AI feature in NeuraLog. Artificial Intelligence Log Feature Extraction is an image processing module that eliminates well log background and grids, giving you a clean, curve-only image. This neural-network-driven feature significantly changes the way we digitize logs.
In this demo we'll capture a couple of AILFE curves and export those digits into a standard LAS. Let's get started.
0:35Running the AILFE processor
Our demo log is loaded into NeuraLog and we run the AILFE processor. AILFE analyzes the image, then color-codes all curves and horizontal grids. If you like, a NeuraLog preference can be set to run AILFE on all logs as they are loaded.
One of the neatest things about using an AILFE image in NeuraLog is that the color-coded features can instantly be turned on and off. Let's look at the well log image in black and white. We can turn off the background and then the horizontal grids, leaving a view of the isolated log curves. And we can add additional curve pixels to color in those dashed lines.
1:16Calibration and depth grids
Let's calibrate a small section of the log track, from 3,800 to 4,000 ft. Notice the thumbnail to the right of the main log, where we have a bird's-eye view to quickly jump anywhere in the log. We can get started with that gamma ray.
Next, we'll add the horizontal depth grids, which provide NeuraLog with precision depth spacing. For a clean, unstretched log we could skip this step, but horizontal gridding is important for longer logs that may have had skew or stretch introduced during scanning. And while NeuraLog has a grid spacing verification function, it's convenient to turn off the image here for a quick check of the grids. They all look good.
2:04Tracing the gamma ray and caliper
Now let's trace the gamma ray. We add the GR type and select one of the solid tracers. Solid X tends to march down the log, only rarely stopping. Solid is a more cautious tracer that will stop when it hits certain levels of uncertainty.
Now let's trace the caliper. It needs a range of 10 to 20 in, and we choose the dot-dash tracer. The dot-dash tracer easily followed the curve, but did make one mistake. We do have another option: we can color in our dashed curve using the color controls. This makes the line appear solid and traceable with a solid tracer.
2:47Reviewing the traced curves
We can now show all curves, turn off all image features, and we're left with the work we did. The two blue vertical bars represent automatic track boundaries that can be adjusted in and out if needed, and the red vertical line is our depth axis. You see our gamma ray in green, and next to it the caliper in blue.
Now let's turn our image back on: the black background, the blue grids and the red curve. With these three colors activated, the AILFE image will always match the original black-and-white log image. No original image is lost with AILFE.
3:25Exporting to LAS
Now let's export our LAS and place it side by side with the original. NeuraLog allows you to zoom all files to the same scale, and then scroll the documents together for a quick check of your digitized file.
That concludes our demo of NeuraLog's AILFE image processor. At Neuralog, we will continue to work to provide you with the latest and greatest AI tools, all while keeping you, the operator, in ultimate control.
