基于低纬度SVM决策树算法的智能手语翻译手套  

作　　者：白旭 郭豆豆 杨学康 蒋丽珍[1] 

机构地区：[1]浙江工商大学信息与电子工程学院,杭州310018

出　　处：《科技视界》2020年第17期33-36,共4页Science & Technology Vision

基　　金：国家自然科学基金项目(61871347);国家级大学生创新创业训练计划支持项目(201911002)。

摘　　要：基于低纬度SVM决策树算法的智能手语翻译手套旨在解决聋哑人与听人之间的交流问题,满足聋哑人在日常生活中的基本社交需求.聋哑人只需戴上该手套,做出相应的手势,手套即可识别出相应的手势并播报语音、同时屏幕显示文本.该智能手语翻译手套采用基于数据手套的手语识别技术,运用低纬度SVM结合决策树算法来进行手势识别,在SVM的两类分类器的基础上加上决策树算法能够实现多类分类器功能,在不影响识别率的情况下极大地提高识别速率.The Sign Language Interpreting Smart Gloves based on SVM Decision Tree At Low Latitude is aimed to bring down the communication barriers between the deaf and dumb and the others,satisfying their demands on social contact in daily lives.Once the deaf and dumb make the specific hand gestures with the smart gloves on,the micro computer of the gloves will analyse the obtained data,then present the corresponding text on a screen,broadcast the synthetic voice according to the identification results at the same time.The sign language interpreting smart gloves combine the gesture identifying technology from digital gloves with low-latitude SVM decision tree algorithm so as to accurately identify the gesture.Without decreasing the identification rate,the identification speed can be dramatically accelerated by the multi-class classifier which can be provided by SVM’s double-class classifier-plus decision tree algorithm.

关 键 词：数据手套 手势识别 低纬度SVM决策树算法 

分 类 号：G06[文化科学] F3[经济管理—产业经济]