Smart wearable robot glasses for human visual augmentation based on human intention and scene understanding

This article focuses on the design and implementation of smart wearable robot glasses for human visual augmentation, which take a role to provide the refined visual recognition result to users of wearing the proposed system. The proposed system consists of a glass-type wearable device with a front looking camera, an eye looking camera, and an earphone, and signal processing units. The scene-analyzing process on the input image ac quired by the front view camera is supported by an eye view camera of monitoring the eye position of user for efficient in formation processing, which is used to catch the user's visual intention and attention in given situations. The recognized results are transformed into the audio information for the user friendly information service without obstructing the users' own visual information gathering and processing, and then the result is transferred into the user earphone finally. This device can be used for the augmentation of human visual capability in various areas, museum, conference, meetings, etc. For the proposed device's feasibility, a series of experiments are performed, and the evaluation results are discussed in detail.