International Journal of Innovative Technology and Exploring Engineering (IJITEE) ISSN: 2278-3075, Volume-9 Issue-3, January 2020 3363 Published By: Blue Eyes Intelligence Engineering & Sciences Publication Retrieval Number: H7062068819/2020©BEIESP DOI: 10.35940/ijitee.H7062.019320 Semantic Image Annotation using Ontology And SPARQL A. Gauthami Latha, Ch. Satyanarayana, Y. Srinivas Abstract: Based on user’s interest or requirements, the search and retrieve images from large scale the databases, the content- based image retrieval (CBIR) technique has become the primary emerging area in research for digital image processing which makes the visual contents to use. Most promising tools for image searching are Google Images and Yahoo Image search. They are used for annotations based on textual of the images. In this, the images are annotated manually with the help of keywords and then the retrieval is carried by using various search methods based on text. Due to this method, the system performance is too low. Therefore, CBIR goal is to construct Image Ontology. The Ontology extracts the relevant images from the database by using low-level features like texture, shape and color. In multimedia technology, the challenging task is to retrieve the relevant images from an image database. For representation, organization and retrieving of images, the searching approaches based on semantic provide effective and efficient results by using image ontology. In this paper, protege software shows us how to create ontology and SPARQL query language provides semantic annotation for images. In addition to this, OntoViz and OntoGraph were used to generate Ontology in a graphical form for the relevant application. Keywords - Image Annotation, Ontology, OntoGraph, OntoViz, OWL, Protégé Semantic, and SPARQL. I. INTRODUCTION In the universe, Web (WWW) is huge database. This discovers files, documents, images etc. for the human. The current search is based on keyword where the machines are lacking in understanding the meaning and relationships among the data. It lacks in semantic structure, where the components scalability and interdependency is maintained . In turn, returns the results for the input query using the hyperlinks among resources and may also retrieve irrelevant information to the user. With the help of hyperlinks, the resources are accessed, where the content is available on the web and here the disadvantage is, the information is in machine readable form which fails in making the machine understand. Therefore, effective search or retrieval of information has becomes highly crucial. In context of given query, the process of extracting relevant results is defined as information retrieval explicitly. We can extract keywords by using various information retrieval techniques. Therefore new search strategies are to be adopted. Semantic based search provides appropriate and relevant results than that of traditional keyword based search. The efficiency of semantic based search depends on Revised Manuscript Received on January 05, 2020 A. Gautami Latha, CSE, SWEC, Hyderabad, India. Dr. Ch. Satyanarayana, CSE, JNTUK University, Kakinada, India. Dr. Y. Srinivas, IT, GITAM University, Vizag, India. the ontology [1]. The process of representation by making use of the properties and relationships among the images are termed as ontology and it is constructed by considering the level of human understanding. Ontology is constructed by making use of the low-level features like texture, color and shape of images which replicates human understanding. Therefore ontology is considered as more useful for retrieval of images that are semantic based. If the feature count varies from high to low, then the ontology construction tends to be inefficient. The, Resource Description Framework (RDF) framework is used to describe and interchange metadata which provide intelligent access among heterogeneous and distributed information and also to construct domain ontology, Web Ontology Language (OWL), is widely used. The two main search approaches for image are: 1. Annotation Based 2. Content Based Annotation based approach is based on keywords or image metadata. [2]-[3]. Images Google Search is one of the examples for this approach. In this the search engine relies on the properties or content of the image. The properties of the image include name/title, creation date, format of the image, resolution, copyright information and so on. Whereas, content based metadata match with the properties of the entities depicted like person, object, etc. The Semantic Web (SW) provides framework which allows data to share and reuse for various applications, enterprises etc and this termed as Resource Description Framework (RDF).  To represent the knowledge resources, RDF is used and to identify the resources, Uniform Resource Identifier (URI) is used.  To represent the resources based on web and SPARQL (Standard Protocol for RDF Query language), RDF Schema is used, to obtain information or data from RDF graphs. These graphs in-turn represented in machine understandable form.  Semantic Search engines and browsers are used for semantic traversal where the agents act as programs for transferring the meaningful data. The paper covers the keyword based annotation structure, and ontology guided reasoning methods for the retrieval of images. Our proposed approach has been implemented using protégé for retrieval using. Images are retrieved using the content description of the images by SPARQL query. II. VISION OF SEMANTIC WEB The idea of Semantic Web (SW) is proposed by Tim Berners Lee in 1996. The vision of Semantic Web is to access data from the