Tuesday, January 8, 2013


Without a doubt DNA Barcoding has modernized biodiversity science and opened doors to countless applications. However, it is just another tool in the toolbox of science and other tools are constantly developed.

Credit Wildtrack
One example for such a new tool is a technique that analyzes images of animal footprints to provide reliable species identification. Two researchers from the Duke University just presented a new method that uses a standardized series of digital images along several different trails to extract variables that provide best classification. The so called footprint identification technique (FIT) can identify species based on an algorithm refined through several training sets drawn from captive animals tracked on different substrates. When tested with free-ranging populations of different tiger species, rhinos, polar bear, cougar and cheetah the method provided not only species but also sex and age class with accuracy greater than 90%.

At this point the method requires the positioning of landmark points on the footprints which is the only drawback as this needs some experience and expertise. This is a quite common problem in biometric analysis and morphometrics. However, the authors discuss options for an automated feature extraction that would complement their development and produce a user-friendly tool to monitor species in the wild using footprints.

No comments:

Post a Comment