Show Details

Detecting Fish

January 31, 2014

An image processing algorithm helps find fish in a sea of data.

Transcript

BOB HIRSHON (host):

Underwater fish detectors. I’m Bob Hirshon and this is Science Update.

(Bruno di Giusti/Wikipedia)

(Bruno di Giusti/Wikipedia)

One consequence of the digital age is that scientists have to sift through increasingly large amounts of data. That’s where University of Victoria computer scientist Alexandra Branzan Albu comes in. For example, she’s helped marine biologists analyze thousands of hours of underwater video footage in search of fish.

ALEXANDRA BRANZAN ALBU (University of Victoria):

And it’s very time consuming for a marine biologist to go over so much video footage in order to study a species of interest.

HIRSHON:

So Albu developed an image processing algorithm that mimics human vision. It recognizes patterns made by the animals’ shape and motion.

ALEXANDRA BRANZAN ALBU (University of Victoria):

You can think about the patterns as the visual signature of a real world object.

HIRSHON:

She’s also working on algorithms to detect less mobile sealife, such as scallops and crabs. I’m Bob Hirshon, for AAAS, the science society.