Gesture action recognition method for Chinese Puyue sign language coding

Abstract

The invention discloses a gesture action recognition method for Chinese Puyue sign language coding, which comprises the following steps of: 1, capturing a Chinese Puyue sign language coding gesture action video by using a camera to obtain a framed RGB image sequence; 2, performing color space conversion on each frame of image and determining a gesture area; 3, determining a fitting circle according to the extracted minimum value of the hand-shaped edge contour point sequence, replacing a palm center area, and then detecting a fingertip and a finger root; 4, extracting the angle and distance characteristics of each contour point by taking the circle center and the vertically downward unit vector as standards; 5, performing mapping transformation on the features by adopting a dynamic time warping algorithm to obtain similar distance features of a certain gesture; and 6, identifying the sign language actions by adopting a plurality of classifiers. The method can effectively identify the Chinese Puyue sign language coding gesture actions, provides a solution for a continuous and large-scale vocabulary Chinese sign language identification system, and is of great significance for popularizing normative Chinese sign languages.

Description

technical field [0001] The invention belongs to the technical field of gesture recognition of Chinese sign language, and in particular relates to a gesture recognition method of Chinese Pule sign language coding, which is mainly used in sign language recognition systems, human-computer interaction and other related fields. Background technique [0002] Gesture action is an important input method of human-computer interaction, and it is applied to various information communication processes with its rich and varied meanings and flexible execution methods. Gesture recognition refers to the use of computers to detect, analyze and explain the gestures performed by people, such as the movement mode and force of arms, wrists, fingers, etc., so as to judge the intention of the movement and provide corresponding services. Gesture recognition technology is widely used in Sign language recognition, rehabilitation training, human-computer interaction, home appliance control, smart cars...

AI Technical Summary

Problems solved by technology

When using the above-mentioned gesture recognition method to recognize Chinese Pule sign language, it cannot achieve a good recognition effect, because the direction correction method described is no longer suitable when using a conventional camera or wearing long sleeves, and the commonly used HOG feature has a relatively small dimension. High, unable to meet the real-time and efficient needs, so it is not suitable for Chinese Pule sign language coded gesture recognition

Images