Abstract
Purpose: :
To optimize monocular depth perception with retinal prostheses through psychophysical analysis, thereby allowing enhanced subject mobility and recognition capability. Since the prosthesis will be implanted in only one eye, patients will not have stereopsis. However, they can use pictorial depth cues to perceive depth in the environment. We used image processing techniques to simulate the physical processing of the retinal prosthesis system, thereby clarifying and optimizing the critical constraints for depth perception.
Methods: :
Image processing techniques in MATLAB were used to pixellate and filter images.
Results: :
In low-resolution images, key monocular depth cues reside in the low spatial frequency image information. Such cues include perspective (parallel lines converge at infinity), and occlusion. The high spatial frequency noise of pixellation edges masks low-frequency monocular depth cues, generating the perception of a flat image. Pixellation also generates false depth cues such as same-size objects (pixels) bordered by parallel lines (pixel edges), indicating a single image depth. Post-pixellation blur removes these false depth cues and unmasks the real depth cues to generate a vivid sense of depth.Nine naïve subjects rated the perceived depth in natural images at varying levels of pixellation and blur. Depth ratings of block pixellated images were compared with block pixellated and Gaussian blurred images of the same resolution. Further study showed that even after recognition of an image feature, a clear difference between the depth ratings of pixellated images and post-pixellation blurred images persisted. Dynamic depth cues were also studied with motion video.
Conclusions: :
Depth perception caused by monocular depth perception cues was found to be impaired when the images were pixellated. The impairment was alleviated with the addition of optimal post-pixellation Gaussian blur. The impairment was found to persist even when object recognition was engaged, indicating the impact of false depth cues.Implications for the design of single-eye retinal prostheses include: (1) Electrode stimulation patterns with optimal field and current spreading may enhance monocular depth perception. (2) Depth perception optimization may allow blind subjects to better avoid obstacles and locate desired objects.
Keywords: depth • image processing • shape, form, contour, object perception