We analyzed the chirp sound and behavior of Teleogryllus emma with observation system, which was consisted of computer, ccd-camera and microphone. Computational methods of wavelet transformation and Self-Organizing Maps (SOM) were utilized to characterized the chirp sound of insect species for automatic counting in this study. Wavelets were initially applied to feature extraction of the chirp sound. Wavelet coefficients were accordingly calculated based on the basis function (e.g., Morlet). The obtained coefficients were subsequently provided to count number of chirps in each song. Sound structure of insect specimens consisted with long chirp and short chirp and the patterns of song were grouped by frequency of long chirp and short chirp. The song patterns of insect specimens were divided by Self-Organizing Map (SOM) that was used number of chirp as input data. Application of computational methods to automatic detection of chirp sound was further discussed for obtaining objective assessment in behavior science.

• Ji Chang Woo(Division of Applied Entomology, National Institute of Agricultural Science and Technology)
• Kim Nam-Jung(Division of Applied Entomology, National Institute of Agricultural Science and Technology)
• Lee Sang-Guei(Division of Applied Entomology, National Institute of Agricultural Science and Technology)
• Hwang-Yong Kim(Division of Applied Entomology, National Institute of Agricultural Science and Technology)
• Choi Hoon-Sik(Hankwang Girl Middle School)
• Song In-Sun(Hankwang Girl Middle School)
• Ko Jee Sun(Division of Applied Entomology, National Institute of Agricultural Science and Technology)