An Intelligent Content-based Image Retrieval System Based on Color, Shape and Spatial Relations

Abstract

Research into content-based multimedia information retrieval is both fascinating and challenging. The use of shape, texture,
and color data is one example of a modern technique. In this study, we provide a hybrid method that has been used to recover
images from around 5,000 photographs; this method takes into account the color, form, and spatial interactions among items
in a picture. Based on our clustering approach and color sensibility, we use a redesigned color scheme and its indexing strategy
to enhance retrieval efficiency. The process of extracting form and spatial connections from objects is akin to seed filling.
When comparing geographical disparities for similarity, a qualitative technique is used. In addition to a user-friendly graphical
user interface (GUI), the system incorporates sketch pictures and relevance feedback to enhance retrieval accuracy. Our results
demonstrate that the system can effectively extract picture data using the suggested method.