21 February 2015

Image editing

It's all well having a sediment core, and it's all well having hundreds, but when you need to work on them you need to somehow extract information from them that you can just keep on a piece of paper or a computer file. You can't dive into the cold store every time you wonder about something! So on board we make a core description, in which we note things like sedimentology and fossil content through depth. And we take pictures. And later on, as one might have noticed, we have them X-rayed in hospital. And it takes 3 to 4 scans to do a complete core section.

Ideally, one would want the description, the picture and the scan all in one figure. But that is a lot of work! So recently, I have spent a lot of time staring at my screen, making snazzy digital versions of the core descriptions made onboard, and adding the core photographs to these logs. And then I had the job of taking the several scans of each section and stitching them together, somehow. Once I had the merged scans I could add them to the core logs too.

My first port of call was David; he's the School's photographer and general graphic man. He uses Photoshop for such jobs. But Photoshop struggles with X-rays! They're a bit fuzzy, and the software just doesn't see the overlap. He showed me how to do it manually, but that's a fair amount of work.

Guy (the lab technician) used some other software for the same purpose, but this software struggled in the same way. And then Juan, my office mate, asked why I didn't use the software they use. Juan, like all my office mates, is a sclerochronologist; a person who uses banded organisms for dating and palaeo-environmental research. Tree rings are the most famous example of annual banding in an organism. But corals have them too, and the ear bones of fish, and statoliths of gastropods, and whatnot. And clams have annual (or finer) bands too. All my office mates use clams. And they take very detailed photographs of their shells, under the microscope, and use the pics for image analysis. And as they need high resolution, they can't do the entire shell in one go. They need to stitch many pics together! So their software was tailor-made for what I needed.

An example of a raw X-ray image; this is some 36 cm of two 100 cm core sections. The circles, btw, are scars of the shear vane, with which we measure shear strength. Shear strength is an indication of compaction; if sediment has had an ice sheet on its head, it tends to get very compacted!

Juan showed me what the software could do. It was magic! So easy! I did all my sections while the sclero people were having lunch. Result! So then I only had to combine the scans with the other data. And when that's done I have all the information I could possibly want about my 88 cores ready at my fingertips! And then the REAL work can start; look for datable material revealed by the scans, dig it out, and send it off for radiocarbon dating. But for now I have clean hands and square eyes! And increasing knowledge on image editing...

A fully integrated core description, with all bells & whistles; notice the X-ray on the left that looks like one picture, but which is 4 pictures merged together

No comments: