Gesture language-voice bidirectional translation system

Abstract

The invention relates to gesture recognition technology and voice recognition technology based on sensing technology and pattern recognition technique, and belongs to the fields of sensing technology, and inertia signal processing and voice processing technology. The invention particularly relates to a gesture language-voice bidirectional translation system. By using the micro-inertial navigationsensor network and the gesture language movement classifier as the core, the invention detects the gesture language movement made by a deaf mute through sensors installed on fingers, arms and head ofthe deaf mute, makes correct recognition and consecutively expresses the represented meaning through a voice and sound production system, so that a normal person can see the gesture language. Meanwhile, the language of a normal person can be translated into specific consecutive gesture language movements and displayed in the form of animation or video, so that the effect that the deaf mute can indirectly hear a normal person is achieved. The system effectively solves the impediment of bidirectional communications between the deaf mute and the normal person, is cheap and portable and has huge social benefit and economic benefit.

Description

technical field [0001] The present invention relates to body movement recognition technology and voice recognition technology based on sensing technology and pattern recognition technology, especially a sign language voice two-way translation system with micro-inertial navigation sensor network and sign language action classifier as the core, which can perform sign language actions Speech expression can also translate speech information into sign language actions for display, and belongs to the field of sensing technology, micro-inertial navigation signal processing, and speech processing technology. Background technique [0002] The pattern recognition of sign language movements is an important link in the sign language translation system. The pattern recognition of the sign language translation systems that are widely used now is based on visual recognition. This recognition method mainly uses multiple cameras to take pictures of sign language actions from various angles o...

AI Technical Summary

Problems solved by technology

However, this pattern recognition based on visual sign language movements is flawed. There are two obvious points: 1. For complex sign language movements, it is difficult to process and recognize them accurately with images; 2. It is very difficult to process and recognize them with images. Easily affected by the environment, such as in low light conditions

Images