During their evolution, insects developed a very efficient visual system which solves complex visual tasks with a minimum of neural computing power. Behaviors like visual course stabilization, obstacle avoidance and goal-directed navigation are performed by these animals with a nervous system weighing only a few milligrams and containing less than a million neurons. Moreover, all this is accomplished under a very wide range of brightness conditions. The great challenge of this interdisciplinary project is the modeling of these highly nonlinear, adaptive biological principles underlying visual orientation behavior and the resource-efficient integration of them into a microelectronic system. Aiming at a vision system with close reference to its biological counterpart, a novel image sensor combining a sampling matrix adapted to the optical needs of panorama vision with local sensitivity adaptation and temporal pre-processing will be developed. The project combines in a unique way the complementary expertise in experimental analysis and modeling of insect vision systems and computationally as well as energy efficient microelectronic embedded hardware design.