The Research And Implementation Of Chinese Sign Language Translation System

This paper focuses on the characteristics of two different style of linguistic:Chinese and sign language,and explained with the knowledge of linguistics.In addition,this paper explores the correspondence between English and sign language as well as the realization form based on Chinese theoretical basis,the research results can be convenient for people to learn and know sign language,and can be widely applied in the field of translation in sign language,and promote us to cognition and study the natural language.From the perspective of linguistics,the "basic word" of Chinese is Chinese characters.Chinese Sign Language also has the concept of—basic word,but the—Chinese Sign Language dictionary uses vocabulary as the unit to describe Sign Language.No mention the word formation of sign language by—basic word.In order to find the word formation rules in sign language vocabulary,this paper measure the relationship between the sign language vocabulary and basic word by DTW algorithm,and proposed the idea of sign language vocabulary encoding.Then the sign language vocabulary is expressed as a combination of the—basic words.In terms of Chinese information processing,the key of the computer understand the meaning of a sentence is whether it could be segmented correctly.And understanding the meaning of the sentence is the presupposition of translated sentence.So in the process of translation of sign language,it is necessary to carry out Chinese sentence segmentation processing.Chinese word segmentation methods presently used to roughly divided into three:based on the maximum matching method,statistical method based on the large-scale corpus,and the method based on understanding.Through to the research of these three segmentation methods,this article summarizes. the advantages and disadvantages of each method.In order to give full play to the advantages of various segmentation methods,For Chinese word segmentation processing this article adopts the double hash dictionary and combined with improved forward matching algorithm,at the same time,Judge sentence while exist of"overlapping ambiguity," and the segmentation algorithm using the Word omni-segmentation algorithm and statistical method to deal with the overlapped ambiguity,For the sentence if there is a combination ambiguous segmentation,this paper proposes a statistical method to find the overlapped ambiguity in Chinese words' segmentation and solve it by DSM algorithm.Combining with the previous study of sign language and Chinese,this paper improves the traditional Chinese treatment method,and studies the language rule of sign language,and the conversion of Chinese language to sign language.This paper realized the function of translation between Chinese and Chinese Sign Language by computer itself,reaching the requirement of the software design specification.