An approximate retina model with cascade structures

Understanding how the retina transforms incoming light signal to spike trains, which convey visual information to the brain, is a key issue for the development of a retinal prosthesis. This paper presents a cascade model of the retina including linear non-separable spatiotemporal filter, static non-linear rectification, sampling and a Poisson spike generation. When a static image is presented to the model, we gain the output image of bipolar cell layer and spike trains of 100 sampled ganglion cells. The simulation results suggest that such a physiological based retina model could be helpful to understand the image processing and encoding mechanism in the retina.