Film festival

A staple of my research is image analysis, which means I spend a lot of time looking at photos and videos of the seafloor. It sometimes amazes me how much you can learn from a single image.

A frame grab from this year's ROV video, showing anemones
and sponges on the paddle wheel of the Portland.
This week, I've started analyzing the ROV video we collected from shipwrecks in Stellwagen Bank National Marine Sanctuary. As you might recall, I lead an interdisciplinary team of archaeologists, ecosystem managers, and videographers on day-trips over the summer and an expedition with telepresence outreach this September. We recorded hours of footage from the steamship Portland and the coal schooners Palmer and Crary. Now, I'm weeding through the footage to glean quantitative data about the biological communities that inhabit the wrecks.

The first step is convert the ROV video to still images. To do this, I watch the video and pause it any time the ROV has a good, clear view of the species on the wreck. Then I save that frame as a photo. I record metadata for each image, such as where on the ship the frame grab was recorded (bow/stern, port/starboard), which direction the wreck surface is facing (vertical/horizontal/upside-down), and whether there's any fishing gear visible in the image. Over time, I accumulate a library of images that each serve as a semi-independent sample. This library of samples will allow me to conduct a quantitative analysis of the wreck community.

It's been like a film festival in my lab this week, with fascinating videos playing one after another. I love data analysis!

Comments