Riconosci le macchie: foto-identificazione del tritone crestato (Triturus carnifex) usando I3S Pattern

In recent decades, amphibian populations were subjected to a progressive worldwide decline. As a result, several studies monitored and studied the dynamics of the population, in an attempt to counteract this trend. Many studies need capture-mark-recapture sessions, in order to estimate key parameters such as the numerical abundance, the density, the phenology, and to collect biological samples and biometric data. In such researches is needed the unique identification of each individual, a fundamental requirement for a correct interpretation of the collected data. Over the past few decades, several techniques have been designed and improved to identify amphibians. Some, such as the toe-clipping, although widely used, have been repeatedly criticized over the years, as potentially harmful to the populations under study. Thanks to the considerable computing power offered today by modern computers, in recent years the photo-identification has arisen as a tool for individual identification. It comprises diverse set of software and techniques that help the operator to identify the captured individuals. We present the results of a medium-term study of a population of the Italian crested newt (Triturus carnifex) in which individuals were photo-identified with sight and with the help of the free software I3S Pattern. The software requires the identification of three reference points to superimpose the images and then proceeds to the automatic extraction of the interest points using the OpenCV version of the SURF algorithm. The software was able to correctly recognize all individuals, both in the test run, both in the subsequent field validation. The efficiency, gratuity and ease of use, make I3S Pattern a useful tool for the individual identification in case of studies involving a large number of animals.