Automatic call recognition of cricket songs using image processing

Abstract

An automatic call recognition (ACR) process is described that uses image processing techniques on spectrogram images to detect constant-frequency cricket calls recorded amidst a background of evening sounds found in a lowland Costa Rican rainforest. This process involves using image blur filters along with binary filters to isolate calling events. The binary filters are used to isolate potential calls from background noise, and the blur filters are used to unite discrete call fragments as a single continuous call. Features of these events, notably the events central frequency, duration, and bandwidth, along with the type of blur filter applied, are used with a Bayesian classifier to make identifications of the different calls. Of the 22 distinct sonotypes (calls presumed to be species specific) recorded in the study site, 17 of them are recorded in high enough numbers to both train and test the classifier. The classifier approaches 100% true-positive accuracy for these 17 sonotypes. The high true-positive accuracy of this process enables its use for monitoring singing crickets in tropical forests.