Feature Analysis of Quantized Histogram Color Features for Content-Based Image Retrieval Based on Laplacian Filter
Fazal Malik, Bin Baharudin · 2012
C olor is most prominent and widely used feature in content-based image retrieval (CBIR). It is most commonly extracted in images by using the histogram. Extraction of features from enhanced image gives good performance in image retrieval. In this paper a CBIR algorithm is proposed for the retrieval of images based on the Laplacian filter for enhancement of image using the statistical quantized color histogram features. The sharpening method using Laplacian filter is used for enhancement of image with significant information for retrieval. The statistical quantized color histogram features are extracted from sharpened grayscale image using different number of quantization bins. The color features are used in similarity measurement of the query image with database images for retrieval of similar images. The retrieval performance of the color features is analyzed for different number of quantization bins in terms of efficiency and accuracy. Experimental results using Corel image database show that the quantized color histogram features of Laplacian sharpened image are robust in retrieval. Keyw ords: Co ntent-Based Image Retrieval (CBIR), sharpened grayscale image, Laplacian filter, color histogram