Image Retrieval using Otsu and Thepade's Sorted Block Truncation Coding Based Color Features

Abstract

Because of the tremendous growth in digital imaging and enhanced communication and storage technology, trillions of pictures are clicked, stored, and exchanged daily. Finding and searching for an image in an enormous collection is becoming challenging. The query by reference image retrieval technique aims to close the semantic gap between the query image and the retrieved images while improving performance. The primary goal of the work proposed here is to develop discriminative and descriptive features of the image with the minimum possible size. Here, the weighted feature fusion-based image retrieval method is proposed using Thepade's Sorted Block Truncation Coding (SBTC) and Otsu Multilevel thresholding (OMT) methods. The technique is tested for two standard datasets with mean square error (MSE) as a distance measure and average retrieval accuracy (ARA) as a performance metric. The technique has contributed to the enhancement of ARA with the small and fixed-size image feature vector. The feature vector generated is much smaller than the image dimension and is used to represent the image for retrieval. Results prove that the proposed technique of SBTC level 8 with 0.5 weight and OMT - 8 thresholds with 0.5 weight feature fusion gives better ARA than other methods studied.