Behavioral modulation by genetic changes garners a special attention nowadays as an effective means of revealing genetic function on the one hand and broadening the scope of in situ monitoring on the other hand. The cGMP-dependent protein kinase was treated to the western flower thrips, Frankliniella occidentalis. Automatic recognition techniques and computational methods were utilized to investigate behavioral changes across photo- and scoto-phases. Movement behaviors are objectively expressed according to parameter extraction and data structure visualization in different light phases. By comapring with the individuals without treatment, activities of treated thrips were changed including decrease in circadian rhythm. Usefulness of automatic monitoring of insect movement in different genetic strains is further discussed for providing useful information on monitoring and diagnosing natural and unntatural genetic disturbances.

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.

In order to achieve the optimized pest control, correct estimation of pest densities is a prerequisite to monitor pest damage and to provide efficient pest management plans. Parameters regarding diffusion (e.g., diffusion constant) and population size (e.g., growth rate) were estimated by using diffusion equation. The time series dispersal data of Whiteflies collected in greenhouse were used for modeling. Cross-correlation analysis was conducted to reveal the range and direction of pest population invasion. Sampling theory was further investigated regarding estimation of densities, and population dynamics of Whiteflies were discussed in two dimensions.

Accurate estimation of pest density is a prerequisite in achieving efficient pest management. An automatic pest detection system with image processing was installed on a robot to recognize brown marmorated stink bug (Halyomorphahalys) on leaves of paprika(Capsicumannuumvar.angulosum). The shape of pest was recognized and subsequently the robot arm was moved toward the leaves to spray pesticides. The detection system was efficient along with increasing population densities increased. The robot with image processing system was useful for estimating population densities in spatial and temporal domain efficiently.

Accurate estimation of insect density is essential for effective pest management. A simple robotics and image processing system were combined to automatically recognize the density of whiteflies. Subsequently the robot arm was utilized to spray the pesticides in the area of infestation in a minimized amount. The estimated densities of samples in the laboratory condition were in accordance with the actual values. The detection system was efficient when the whitefly densities were at medium to high levels. The results of the present study indicate that the robotic and image processing integration system described here would be useful for evaluating the population dynamics.