Abstract
In biological vision, the retina is organized into a space-variant sampling structure including a high-resolution, small central fovea, and a periphery whose resolutionlinearly decreases in steps. By this characteristic, it able to centralize and distribute the processing loads in the earliest stage vision. Therefore it is essential for smart sensing to integrate functions of spatially variant flexible sampling control system onto computational sensor. The integration on the sensor focal plane results in enhancement of performance of an image sensor.
