Group movements of insects are bases for unravelling origin of social behavior of animals and are important in both theoretical (e.g., evolution) and practical (monitoriing) aspects. Automatic recognition and effective computational methods were developed for characterizing multi-individual interactions in laboratory conditions. Movements of Drosophila species in different genetic strains were continuously observed across days. Characteristic behaviors are objectively expressed based on parameter extraction and data structure visualization. Group activities, including aggregation, inter-individual interactions and arena positioning were objectively characterized in different photo- and scoto-phases according to machine-learning and spatio-termporal patterning techniques. Individual-group relationships are presented regarding how individual movements would contribute to formulating group activities. Usefulness of automatic monitoring of insect group movement is further discussed for a basis for genetic functioning in behavioral aspect.

• Tae-Soo Chon(Research Institute of Computer, Information and Communication, Pusan National University, Ecology and Future Research Institute)
• Yong-Hyeok Jang(Ecology and Future Research Institute)
• Hye-Won Kim(Ecology and Future Research Institute, Deptartment. of Electrical and Electronics Engineering, Pusan National University)
• Chunlei Xia(Research Institute of Computer, Information and Communication, Pusan National University, Ecology and Future Research Institute)