Font Size: a A A

A Comparative Study On The Effects Of The Semantic Annotation Of Dunhuang Mural Digital Images From The Perspective Of Users

Posted on:2019-05-17Degree:MasterType:Thesis
Country:ChinaCandidate:J W GuoFull Text:PDF
GTID:2405330545483374Subject:Management Science and Engineering
Abstract/Summary:PDF Full Text Request
Mogao Grottoes in Dunhuang is a treasure in Chinese cultural heritage.It has extremely high research value and artistic value.With the rise of Dunhuang studies and the development of digital humanities,more and more art resources such as caves,murals and suicides have been collected.After the emergence of large-scale digital projects such as digitalization and imagery,and "digital Dunhuang",a large number of mural digital image resources have been accumulated.In order to effectively store,retrieve,and use such digital resources,more and more scholars have begun to carry out deep semantic annotation of Dunhuang murals.At present,the mainstream image semantic tagging method is based on metadata tagging.However,because of the semantic gap between high-level semantic information and underlying semantic information,the metadata method has many deficiencies in annotating the digital image of a mural,and the mural image cannot be revealed well in terms of their content and semantic connotation.In order to make up for the defects of metadata annotation,the method of annotation of digital image structure began to emerge.The SDFDI has a more accurate modeling of image content semantics.At the same time,the use of thematic vocabularies and image metadata,together.It is to achieve high-level semantic expression of images.And according to it,a set of Dunhuang digital image annotation system for experimental mural annotation work was exploited.One of the main purposes of the deep semantic indexing of Dunhuang mural digital images is to allow users to better understand and appreciate this historical and cultural heritage.Few scholars are currently concerned about this.The purpose of this study is to examine the extent to which the SDFDI approach can enhance the user’s understanding of murals and the level of understanding that can be used to enhance the user’s understanding,and compare the annotation based on the SDFDI labeling system and the labeling effects of traditional digital museum labeling materials.The Effect of SDFDI Marking Dunhuang Mural Digital imagery.The first chapter mainly introduces the research background of Dunhuang digital image semantic annotation,reviews the current research status in this field,and expounds the research methods and contents of this research.It leads several sub-problems in this paper,and then introduces the research ideas and main contributions of this paper..In the second chapter,we sum up the predecessors’ research summarization to sum up the methods of semantic annotation of imagery;secondly,the application of these semantic annotation methods in Dunhuang fresco digital image annotation;and finally,the common method of semantic annotation of Dunhuang fresco digital images."One-Symbol" annotation method of "metadata" and SDFDI hierarchical semantic description model.The third chapter,the first part describes the user’s basic concept of art image recognition,which introduces in detail the theory involved in the study of user image recognition,which is mainly the Panofsky three-level model widely accepted by everyone in the field of art understanding.Referential theory;and the important meaning construction theory in the process of user image cognition and imagery.The second part describes the experiment design ideas based on user’s image cognition from this study.It mainly includes how to obtain the user’s understanding of information needs of murals,how to extract user information needs into questions,and classify problems,and finally explain how the research is done.The labeling effect of the two labeling methods is evaluated based on the user’s perspective.The fourth chapter mainly describes the experimental process of this paper.First,it edits the user questionnaire according to the questions extracted from users’interviews.The questionnaire is used to determine the user’s understanding of murals,finally,users’ experiments were conducted to issue questionnaires and to recover questionnaires,statistical data,measurement reliability and validity.The fifth chapter is used to analyze statistical data,compare the labeling effects of the two labeling systems,and discuss specific issues in the questionnaire(for example,whether different labeling methods can better solve the problem of high user attention).The sixth chapter describes the user’s concerns and discusses the advantages and disadvantages of the two labeling methods from the perspectives of different types of murals and different frequency of the same type of murals.In the "Medicine Buddha illustration",the labeling effect based on the SDFDI labeling system is much better than the "marking materials" created by Dunhuang Academy.The last chapter is mainly a summary of all the research work of this article,including the contributions and deficiencies of this research,as well as the outlook for the possible future research work.
Keywords/Search Tags:Image semantic annotations, Semantic annotation effect, SDFDI, Semantic annotation based on Metadata, Dunhuang frescoes
PDF Full Text Request
Related items