伺服器輔助之可攜式手形辨識系統

Server-Aided Recognition of Finger Shapes in a Portable System

林政緯(Cheng-Wei Lin)

指導教授 ： 黃士殷

元智大學/資訊學院/資訊工程學系/碩士(2008年)

https://doi.org/10.6838/YZU.2008.00234

摘要

喑啞人士無法以言語表達內心想法，因此手語是他們最常使用之溝通方式。
本論文提出一個可攜式手語通訊系統，讓喑啞人士利用簡便的操作輸入手語，
經由無線網路傳輸至遠端辨識伺服器即時處理之後，輸出語音或文字與一般人溝通。
傳統手勢辨識之研究方法主要可分為：影像處理與感測器手套，
其中影像處理的步驟主要為膚色偵測以及動作偵測，
因此必須處理大量的資料，造成耗費大量頻寬與系統資源，
對於可攜式系統電力之損耗過大，使得系統運作效率不彰；
而感測器手套價格昂貴且造成使用者不便，難以符合一般需求。
本論文研究手語手勢之特徵，擷取出手語辭彙之關鍵訊框，
並且利用關鍵訊框改進手勢辨識影像處理之流程，
以降低運算成本、減少無線網路頻寬耗費以及提升系統可用性。

關鍵字

台灣手語 ； 手指辨識 ； 分散式辨識

並列摘要

Sign language recognition had been developed for many years, but none of these methods consider about the convinence and availability for
deaf community. We propose a client-server system: MuteTone, which is designed for easy to use with low-cost equipments.
MuteTone consists of a hand-held device client and a computer server, the hand-held device is used to capture the user's gestures
, transmit data to the computer server through wireless network, and receive the result of a recognition for multi-lingual sound or text output.
In this system, deaf people with a MuteTone client could communicate with normal people more easily and efficiently at many places.
This paper implemented the system and examined about the power consumption of the client, the usage of wireless network bandwidth and
a set of recognitions in Taiwan sign language.

並列關鍵字

Taiwan sign language ； finger recognition ； distributed recognition

參考文獻

[1] 王曉書, 天使之翼 - 曉書手語, 時報文化出版企業股份有限公司, 1st ed., 2003.

[6] G. Nagy, “Interactive, mobile, distributed pattern recognition,” ICIAP, 2005.

[7] J. J. LaViola Jr., “A survey of hand posture and gesture recognition techniques

vision-based taiwanese sign language interpretation,” ACII 2005 LNCS 3784,

[12] J. Park and Y.-L. Yoon, “Led-glove based interactions in multi-modal displays