This article presents a new design of the visual prosthesis based on the optical stimulation. The aim of this work is to seek an architecture of image acquisition, by using a simple treatment and without recourse to extensive treatment of a processor, and to provide an outlook on the image quality acquired by the patient. To do so, we have proposed two architectures of image processing, which have tested with different images. Simulation results demonstrate that the patient can recognize objects and the projection of targets, read letters (scale 20/200 and 20/100) and identify faces and scenes.