3D Shape Acquisition System Dedicated to a Visual Intracortical Stimulator

We present in this paper the architecture of a stereoscopic-based range finder system. The proposed system is dedicated to a visual intracortical stimulator, which is intended to create artificial vision for people suffering from visual blindness. The 3D analysis of the environment is required allowing a better autonomy to the patient. Our system consists of an emitter projecting light patterns from an infrared light source reflecting on a micro-mirrors matrix. Images from the illuminated scene are captured by a camera. The three-dimensional range is reconstructed from the distortions in the reflected and captured image. Our approach consists of taking advantage of a high-speed camera and adapting the Gray codification strategy to dynamic scenes. Tests have been made onto a moving plane surface leading to an error of 1.2 centimeters over a video sequence of two hundred range images.