Abstract
Animals are able to perceive detailed visual features in natural scene, which often exhibits high contrast ratio across the field of view. Although being a fundamental feature of the visual system, the circuit mechanism of the high dynamic range is not well understood. To address the problem, we developed a computational model of the local neuron (M6LN) circuit in the sixth layers of the medulla, a neuropil in the early visual area. A recent study revealed that the M6LNs have diverse coverage patterns in the medulla and suggested that M6LNs may take part in tuning the dynamic range of the visual system. 1 We systematically investigated the correlation between the coverage patterns of M6LNs and ability of dynamic range tuning. We tested our model with images of natural scene and found that the combination of N6LNs with specific coverage patterns led to the optimal results. Under this circuit configuration, the visual system is able to process images with high contrast ratio in a way similar to the high-dynamic rang (HDR) technique used in modern cameras.