Skip navigation
Researchers Work on an Image Database of Insects

 

Narrator: This is Science Today. Have you ever found an insect in your home and then looked up images on the Internet to determine just what it was? Many computer users are doing just that and now engineers and scientists at the University of California, Riverside are working to make that task easier by creating an image database to classify insects. But computer scientist Eamonn Keogh says the purpose of this database goes beyond simply satisfying curiosity.

Keogh: There are applications for commercial farmers, for example. About ten years ago, a new insect appeared and in a couple of years, it caused about forty million dollars of damage to the wine culture in Southern California. So, recognizing an insect can be very, very important and it's even difficult for a specialist like farmers.

Narrator: Classifying insects for an image database is difficult because they are so small and there are so many different kinds.

Keogh: So, we're doing various computer science and engineering things and pulling out the color features, the texture features. It's definitely not a problem that's going to be solved completely in the next decade, but we're making incremental progress all the time.

Narrator: For Science Today, I'm Larissa Branin.