A Wavelet-Based Model for Acoustic Identification of Bat Species
Christopher C. Bay
Dr. Jason E. Miller, Faculty Mentor
Acoustic identification of bat species is important to field biologists and land managers who survey bat populations. Qualitative identification requires highly skilled and experienced researchers and yet still remains very subjective. A successful quantitative identification model would be more accurate and usable for field biologists. Current quantitative models, while having high success rates, often fail to distinguish between similar species. We are trying to develop a quantitative wavelet-based model that will improve on the accuracy rates of these quantitative models. Working primarily with the mathematical software packages Matlab and Octave, we integrated wavelets and cluster analysis. The results of this research will be presented along with possible directions for future work.
Keywords: bats, acoustic, identification, wavelets, Matlab, cluster analysis
Presentation Type: Oral Paper
Location: VH 1232