Effective Image Representation using Double Colour Histograms for Content-Based Image Retrieval

Image representation is critical to the successful realisation of Content Based Retrieval (CBIR) systems. The choice features represent image affects retrieval performance. Nowadays, databases are heterogeneous, and different feature types can be used for appropriate descriptions. This paper proposes an CBIR that combines Stacked Colour Histogram (SCH) Conventional (CCH) improve precision. presented technique designed capture colour texture information image. properties represented by CCH SCH. weighted similarity measure estimate proportion values in task. novel descriptor has been widely tested on four standard datasets, namely Batik Coil100, Corel10K Outext. Batik, Coil100 Outext assess discrimination. discrimination heterogeneous images. Experimental results comparisons with SCH, CMTH, MTH, TCM, CTM NRFUCTM demonstrate proposed superior