Infrared Image Processing Strategy For Enlarging The Working Environment Of Retinal Prosthesis

As the aggravating of global aging population,the number of visual disability has risen year by year,and blindness has become one of the most serious disabilities that threaten the quality of human life.Among the massive blinding diseases,there is no effective cure for blinding diseases caused by photoreceptor cell degeneration of the retina,such as age-related macular degeneration and retinitis pigmentosa.The emergence of retinal prosthesis offers a possibility for visual function repair.At present,most of the existing retinal prosthesis receive the reflected light from the surface of the object through an external camera to obtain the external scene information,and then encode and integrate the image information through the image processing unit.The micro-current stimulator,which is transmitted to the body wirelessly,drives the electrode array to stimulate the visual nerve pathway and induce the halucinatory vision.However,when there is insufficient or no light,the camera will not be able to capture the image information properly,so the retinal prosthesis will not work effectively.A simple and valid solution is to use a dual-mode camera(which automatically switches between visible and infrared mode depending on the intens ity of the external light is strong or weak)instead of the ordinary camera to expand the working environment of the retinal prosthesis.In addition,the quality of infrared imaging in infrared mode is far less than that of natural light in visible light mode,so the quality of infrared imaging in infrared mode is much worse than that in visible mode.This study designed a psychophysics experiment under simulated prosthesis vision to explore the ability of subjects to complete visual tasks in two imaging modes.The results showed that the ability of subjects to complete visual tasks in infrared mode was significantly lower than that in visible light mode.Since the image processing unit alows image processing algorithms to potentially improve the visual perception of prosthetic implants,and considering its imaging characteristics(low contrast,low signal-to-noise ratio(SNR))in infrared light mode and the features of prosthesis vision(limited gray level,color,texture,etc.)simultaneously,the infrared image enhanceme nt algorithm suitable for retinal prosthesis is explored,and the validity of the proposed method is verified under simulated prosthesis vision.In this paper,an infrared image enhancement strategy based on adaptive platform histogram equalization is proposed.With a mature and effective adaptive platform histogram equalization algorithm,the gray scale of infrared image is stretched and the contrast of the image can be improved.At the same time,the guidance filte r ing algorithm is used to remove the background noise of the image,and considering the edge profile information is very necessary for the prosthetic implant to identify the object features,therefore,On the basis of guided filtering,edge detection algorithm is used to extract the edge of the image.The edge contour is superimposed on the image after the gray scale is stretched to realize the edge enhancement of the image.The purpose of this study was to expand the working environment of retinal prosthesis and to improve the prosthetic implants' ability to perform visual tasks in infrared mode by means of infrared image processing strategy.With the help of the simulation prosthesis vision under the psychophysics experimental platform,firstly,the subjects in the dual-mode camera can continue to work without visible light,and can sense the conversion between the two camera modes.Secondly,the ability to complete visual tasks in infrared mode was significantly inferior to that in visible mode.This suggests that we need a certain infrared image processing strategy to improve the performance of object recognition in prosthetic vision.Finally,the image processing strategy proposed in this paper significantly improves the ability of participants to complete visual tasks in infrared mode.The above experimental results indicate that the proposed scheme can provide a reliable reference for the future design of retinal prosthesis.