Font Size: a A A

Retrieval Of Ancient Chinese Character Images Based On Hesitant Fuzzy Features

Posted on:2020-08-13Degree:MasterType:Thesis
Country:ChinaCandidate:Y L ChaiFull Text:PDF
GTID:2405330596985311Subject:Computer Science and Technology
Abstract/Summary:PDF Full Text Request
Because of the complex structures and the various changes of glyphs of ancient Chinese characters,as well as the deterioration of their glyphs caused by time,the traditional image feature extraction and retrieval methods have a poor performance in dealing with ancient Chinese characters images and there are many problems to be solved.In this thesis,an ancient Chinese character image retrieval method based on hesitant fuzzy features is proposed through using the advantages of hesitation fuzzy set theory in dealing with multiple features and multiple memberships in the integration of multiple structural features and statistical features.Based on the analysis and induction of the characteristics of ancient Chinese characters,the key parts in the ancient Chinese character image retrieval system were studied and designed.Firstly,aiming at the images of ancient Chinese characters and the characteristics of their existed mediums,the image preprocessing method including binarization,denoising,segmentation and skeleton extraction was designed to lay the foundation for further feature extraction.Secondly,the retrieval features of ancient Chinese character images were selected and the evaluation attributes as well as the related evaluation indicators are constructed.The corresponding membership function was defined for each evaluation attributes.On this basis,the hesitant fuzzy sets of ancient Chinese character image retrieval were constructed.Finally,the similarity measure theory of hesitant fuzzy sets was adopted to convert the similarity measure between the images of the two ancient Chinese characters into the similarity measure between the hesitant fuzzy sets composed of the ancient Chinese characters for realizing the orderly output of the result images of ancient Chinese character retrieval.An ancient Chinese character image retrieval system based on hesitant fuzzy sets is realized according to the method of this thesis.The image proofs in the ancient books "Siku Quanshu" were used for experimental analysis.The experimental results show that the proposed method has a certain degree of improvement in precision,recall and F value and it can be better adapted to the characteristics of ancient Chinese characters.
Keywords/Search Tags:Ancient Chinese character, Image retrieval, Structure features, Statistical features, Hesitation fuzzy set, similarity
PDF Full Text Request
Related items