Fixed Representative Colors Feature Extraction Algorithm for Moving Picture Experts Group-7 Dominant Color Descriptor

Abstract

Problem statement: MPEG-7 has defined a set of color descriptors for the multimedia contents; one of these descriptors is the Dominant Color Descriptor (DCD). The DCD adopted the Generalized Lloyd Algorithm (GLA) as a feature extraction technique. However, using this algorithm in image retrieval systems may produce perceptually irrelevant images in the ranked results. Approach: In this study, we proposed a new technique for the extraction of the DCD called Fixed Representative Colors Feature Extraction Algorithm (FRCFE). This technique is based on converting the original image colors to a set of predefined colors in the RGB color space. Also, we developed an image retrieval system to measure the effectives of this technique. Results: Experimental results revealed that FRCFE provides better accurate and perceptually relevant image retrieval for the MPEG-7 DCD than GLA in the chosen dataset. Conclusion: Mapping the original image colors to a set of predefined colors preserved the main features of the image and can produce more accurate results when used in image retrieval systems.