A Web-Based Sign Language Translator Using 3D Video Processing
A Web-Based Sign Language Translator Using 3D Video Processing
复制标题
使用 3D 视频处理的基于网络的手语翻译器
DOI：
10.1109/nbis.2011.60
复制
发表时间：
2011
期刊：
影响因子：
--
通讯作者：
Stephen Lau
中科院分区：
文献类型：
--
作者：
K. F. Li;Kylee Lothrop;Ethan Gill;Stephen Lau
北美和世界其他地区的听障人士使用美国手语(ASL)来补充土著手语。利用3D视频处理技术，设计并开发了一个概念验证的ASL翻译器。预见到它作为基于Web的应用程序的潜力，翻译人员必须有一个便携式输入设备来捕获手势，并且必须保持较低的成本。最近推出的Xbox Kinect是一款多功能的手势输入设备，也符合低成本的要求。分析由Kinect捕获的用户关节的3D数据，并将其与预先记录的标志库进行匹配。然后，匹配的手势被转录成单词或短语，并输出到合适的用户界面。实现的原型在有限的词汇表上具有极高的精确度。使用网络和服务器对预先记录的手势进行存档，并通过动作捕捉设备处理记录的手势，有许多潜在的应用。该翻译器可以作为听障人士交流的辅助工具，也可以作为想要学习手语的人的教学工具。
The American Sign Language (ASL) is used by hearing-impaired people in North America, as well as in other parts of the world to supplement indigenous sign language. A proof-of-concept ASL Translator has been designed and developed using 3D video processing techniques. Foreseeing its potential as a Web-based application, the Translator must have a portable input device to capture gestures and its cost must be kept low. The recently introduced Xbox Kinect is a versatile gesture input device and fits the low-cost requirement as well. 3D data of the joints of a user captured by the Kinect are analyzed and matched to a library of pre-recorded signs. The matched signs are then transcribed to word or phrase, and output to a suitable user interface. The implemented prototype works with excellent accuracy for a limited vocabulary. Using the Web and a server to archive the pre-recorded signs and to process recorded gesture via a motion capture device, there are many potential applications. The Translator can be utilized as an assistive tool for the hearing impaired to communicate or as a teaching tool for those who want to learn the sign language.