Recently, patterning and analyzing complex data of behaviors of animals in response to external stimuli such as toxic chemicals has become focus of attentions. In this paper, an energy minimization model to extract the features of response behavior of chironomids under toxic treatment is proposed, which is applied on the image of velocity vectors. The model is based on the improved active contour model and the variation of the energy values produced by the evolving active contour. We attempt to implement an adaptive computational method to characterize the changes in response behaviors of chironomids after treatment with an insecticide, diazinon. Active contour is formed around each collection of velocities to gradually evolve to find the optimal boundaries of velocity collections through processes of energy minimization. The energy minimization model effectively reveals characteristic patterns of behavior for the treatment versus no treatment, and identifies changes in behavioral states as the time progressed.