Head carried stereo vision hand gesture identifying device

Abstract

The present invention provides a head carried stereo vision hand gesture identifying device and belongs to the field of computer vision and man-computer interaction technology. The head carried stereo vision hand gesture identifying device consists of four parts: stereo vision imaging head, stereo vision board, head carried display and worn computer. The stereo vision imaging head obtains the hand grey image of the wearer; the stereo vision board calculates the real-time the dense parallax map and transmits the dense parallax map and the grey image to the worn computer; and the worn computer completes hand tracking and hand gesture identifying operation and showing the calculation results in the head carried display. The present invention can realize the hand tracking and hand gesture identifying under worn computer environment, complete hand gesture mouse, hand language identification, hand written character identification, etc. and intelligent man-computer interface based on hand gesture.

Description

technical field [0001] The invention is a head-mounted stereo vision gesture recognition device, which belongs to the technical field of computer vision and human-computer interaction, and is used for gesture interaction of a wearable computer. Background technique [0002] Various input devices of wearable computers, such as keyboards, mice, and writing boards, are all worn on people's bodies, such as waists, chests, shoulders, arms and other positions. These input devices are relatively cumbersome for the wearer, so people are vigorously developing gesture recognition technology to solve the natural interaction between humans and computers. At present, various devices and technologies have been used to solve the problem of gesture recognition under wearable conditions, which can be mainly divided into two categories. The first type is a gesture recognition device based on a data glove, such as "Tinmith-Hand" in Australia. However, this type of device is bulky and require...

AI Technical Summary

Problems solved by technology

The main disadvantages of existing devices of this type are: (1) a single camera is used to obtain limited information and is easily affected by ambient light and complex backgrounds; (2) the camera is fixed in direction and can only face forward, and the wearer must Lifting the hand is prone to fatigue and lack of flexibility; (3) Skin color information is mainly used for hand tracking and gesture recognition. Skin color is easily affected by lighting conditions. As the wearer walks, lighting conditions will change, thus Leading to the failure of tracking and recognition; in addition, when an object similar to skin color appears in the camera's field of view, it will cause ambiguity and cannot be distinguished

Images