Enhanced autonomous hands-free control in electronic visual aids

Inventors

Watola, David • Cormier, Jay E. • Kim, Brian

Assignees

Eyedaptic Inc

Abstract

Adaptive Control Systems and methods are provided for an electronic visual aid, which may be wearable, hand held or fixed mount, including autonomous hands free control integrated with artificial intelligence tunable to user preferences and environmental conditions.

Core Innovation

The invention provides autonomous enhancements to a video output of a visual aid device worn by a user. It obtains real-time images of a scene with a camera of the visual aid device and evaluates a motion component of the visual aid device to determine whether the user is engaged in a concentration state, a scanning movement state, or an animated movement state.

When the user is engaged in the animated movement state, the invention presents video images on a display without additional enhancements or adjustments applied to the real-time images. When the user is engaged in the scanning movement state, it locates structures of interest from the real-time images, applies visual enhancements to at least the structures of interest, and presents video images including the visually enhanced structures of interest on the display.

For low-vision enhanced vision, the invention captures real-time video images with a camera of an electronic device and transmits the real-time video images from the electronic device to a visual aid eyeglasses device separate from the electronic device. It identifies a subset of each real-time video image corresponding to a focus of attention of the camera, extracts structures of interest within the subset, identifies at least one group of structures of interest organized in a row or column, determines a dimension of the group, and adjusts a magnification of the group to match a preferred structure dimension for display on the eyeglasses device.

The extracted structures of interest can be text-like structures, facial features, or other structures such as letters, numerals, pictograms, glyphs, or icons. The vision-enhancement approach can use extraction of maximally stable extremal regions from the subset of each real-time video image, and then generate enhanced video images displaying the visually enhanced structures of interest on one or more displays of the visual aid eyeglasses device.

Claims Coverage

The independent claims include three distinct inventive methods: autonomous enhancement of a wearable visual aid video stream based on motion-state classification, enhanced vision using transmitted real-time video with focus-of-attention subset extraction and row/column organization with magnification matching, and enhanced vision using transmitted real-time video where visual enhancements are applied to extracted structures of interest for display on eyeglasses displays. Across these independent claims, the inventive features center on real-time video acquisition, motion-state or focus-of-attention determination, extraction of structures of interest, and display of visually enhanced results.

Across the independent claims, the coverage centers on obtaining or transmitting real-time video images, identifying either a motion engagement state or a focus-of-attention subset, extracting structures of interest, and presenting enhanced video on a visual aid device or eyeglasses display. The row/column grouping and magnification matching of a preferred structure dimension is specifically claimed for enhanced vision, while visual enhancements applied to extracted structures of interest are claimed in the other approaches.

Stated Advantages

Documented Applications

No documented applications found