A Content-Based Indexing System for Image Retrieval

Background/Objective: The aim of this research paper is to create an image indexing system by identifying and explaining image features. In this research work we are developing an image indexing algorithm. Method/Analysis: From the previous researches we select several features that can be considered suitable and can be implemented with the help of Global feature – Boolean Edge Density, Edge Density, Color Sigma, Edge Direction, Color Average and Region feature – Moment Invariant, Grey Level, Region Area. Finding: We identified best combinations for different image data set. The experiments show that region based features increase the performance of image retrieval. Application/Improvement: While comparing the two image features, i.e. global features are less substantial than the region features.