Researchers have devised a software system that can analyze a spider web and find out which species spun it. It's intended to aid scientists in quickly identifying the prevalence of certain species for the purposes of biodiversity conservation. The biometric system uses a database of 150 previously existing cobweb photographs and subjects them to multiple techniques that isolate the web's shape. Since the photos weren't taken specifically for this research, the study was designed to cut through the "noise" and background information that would occur in most images.
The center of the web contains the most salient information, with the outer reaches also helping to contribute; one combination of techniques managed to identify the spider to a 99.96 percent degree of accuracy. The team at Spain's Universidad de Las Palmas de Gran Canaria only used four species in the study, but with such strong results from the initial research, it's easy to imagine how its principles might be expanded to other spiders in the future.