Time Comparison of Various Feature Extraction of Content Based Image Retrieval

Abstract--Recently there has been tremendous increase in usage of multimedia contents like images and videos. So there has been a collection of varied image databases and video databases in digital format. These databases are more widespread and increasing larger. A system has been implemented for efficient and effective search engine and the system is referred as content-based image retrieval (CBIR) systems. Semantic gap is the drawback of CBIR systems. CBIR systems cannot extract and perform poorly in extracting high-level (semantic) features which include objects, feelings and their meanings. Semantic Gap can be reduced in CBIR system by using User Feedback. It is also referred as Relevance Feedback. This paper presents Content Based Image Retrieval (CBIR) system that uses multiple feature fusion to retrieve images and time is compared for each and combination of various features and also for improving the retrieval results in terms of its accuracy relevance feedback is suggested. The features like color and shape are used. The color histogram is used to extract color feature and edge detection is used for shape extraction. This paper also presents the time comparison for different CBIR systems. The experimental results show that the enhanced CBIR system is better in terms of system evaluation parameters like recall and precision