Image Processing And Coding Strategy For Visual Prostheses

Thousands of blind people suffer because of congenital reason or prevalent eye diseases. With the development of biomedical engineering, computer science and bionics, visual prosthesis become an available approach for the blind to restore partial or complete sight. Current microelectrode technology allows the possibility to implant microelectrodes in visual pathway of human, and experiments of implant abroad have improved it.Visual prostheses differentiate according to the different microelectrodes location, in visual cortex, optic nerves and retina. Structures of them are similar though, all need the image processing part to process the image and decode, RF to transmit the image information to simulator implanted internal which send electrical current to microelectrodes.In this manuscript, considering the features of phosphenes array, we studied real-time image processing strategy for visual prostheses and presented the processing results based on TMS320DM642 DSP(Digital Signal Processor) platform. According to the experiment results about relationship between features of electric current pulses and brightness of phosphenes done by research teams abroad, we built a image information-electrical simulation model, using which to simulate the pathway from image acquisition to forming electrical current.