Image redistribution and augmentation for visual impairment
Publication Date: 2019-Jan-11
The IP.com Prior Art Database
Image Redistribution and Augmentation for Visual ImpairmentDisclosed is a general-purpose method that enables the adaption of a screen’s contents to accommodate visual impairments, principally macular degeneration and tunnel vision. The invention proposes a learning algorithm, which can be manually adjusted, to adapt electronic displays such that the entire display is visible to a suffer of degenerative vision conditions. Current support technologies focus on magnification for image viewing and text to speech software for reading. These technologies specifically address only a single issue. Additionally, in the magnification case, the solution is potentially insufficient in the later stages of macular degeneration. The invention provides an all-in-one solution by mapping the screen’s contents to the periphery or center of the screen, dependent on the condition, hence making the screen's contents visible to the suffers of these conditions. Macular degeneration is estimated to affect 6.2 million of people worldwide and is most commonly found to occur in those over the age of 50. The condition presents with the loss, and potential warping, of a person's central vision. There are two varieties: wet, which accounts for approximately 10% of cases and has potential treatment options in laser surgery; and dry, which accounts for 90% of cases and for which there is currently no cure. Tunnel vision is, in effect, the opposite condition. It is depicted by the loss of a person's peripheral vision. This condition has a large variety of causes, from temporary causes such as panic attacks, to progressive, incurable, diseases like Retinitis Pigmentosa, which is estimated to affect 1 in 4,000 people. These impairments can affect a person’s ability to work, read and write, drive and recognize faces. Making use of the user's responses whilst viewing an Amsler grid, the software will learn the inverse distortion needed to minimise the diseases effects on vision. This inverse distortion will then be applied to the screen based on the user's gaze. This learning process can be recalibrated whenever it is felt useful. The degree of fish-eye lens (FEL)-like transformation will be controlled with a simple slider bar that the user can set manually. With large enough datasets, a predicted disease progression could be used to automatically increase the FEL-like transformation with time. Similarly, the precise anti-distortion mapping used could be predicted reducing the number of recalibrations needed. Making use of gaze tracking and a model of pathologies of human vision, the software proposed transforms the image displayed by a monitor (or multiple monitors) in real time such that the distortion experienced by a user with a visual pathology is minimised. Current approaches to this problem offer homogenous expansion of a portion of the screen but, in contrast to the approach presented here, do not offer a transformation of the screen tailored to benefit a particular use...