Depth filter for visual prostheses
The present invention is a depth sensing visual prosthesis. By depth sensing objects within an image, items located at a certain depths from the user can be removed to declutter an image. The resulting image presents only objects within a range of depth from the visual prosthesis user. Depth sensing can be accomplished in various ways including structured light, time of flight, or preferably, a stereo camera pair.
1. A visual prosthesis comprising:
a camera for producing a visual image;
a video processor receiving data from the camera and including a digital image processing system that creates a disparity map, detects and digitally removes portions of the visual image that represent information outside of a preset range of depth from the visual prosthesis user based on the disparity map, and producing stimulation signals based on a portion of the visual image found within the range of depths from the visual prosthesis user;
user controls on the video processor adapted to control depth filtering modes and a band pass filter with adjustable width to control the range of depths visible to the user;
an implantable neurostimulator configured to stimulate visual neural tissue in accordance to the stimulation signals.
2. The visual prosthesis according to claim 1 , wherein a second camera with the camera for producing a visual image form stereo cameras and provide depth information to the image processing system.
3. The visual prosthesis according to claim 2 , wherein the digital image processing system applies the disparity map to data from the stereo cameras to determine depth information.
4. The visual prosthesis according to claim 3 , wherein the disparity map is developed using a block matching algorithm.
5. The visual prosthesis according to claim 1 , wherein the camera is configured to be mounted on, and move with, a visual prosthesis user's head.
6. The visual prosthesis according to claim 1 , further comprising an inertial sensor to provide orientation information of the camera to the digital image processing system.
7. The visual prosthesis according to claim 1 , wherein the digital image processing system removes unwanted image information with a mask.
8. The visual prosthesis according to claim 7 , wherein the mask is refined using open and close morphology operations.
9. The visual prosthesis according to claim 1 , wherein the stimulation signals are based on the presence or absence of an object within the range of depth, regardless of the luminance level of the object.