A New Palette Histogram Similarity Measure For MPEG-7 Dominant Color Descriptor (TA-P4)

Author(s) : Lai-Man Po (City University of Hong Kong, Hong Kong) Ka-Man Wong (City University of Hong Kong, Hong Kong)

Abstract : The MPEG-7 dominant color descriptor (DCD) is an effective, compact, and intuitive representation of salient colors in an image region and is very useful for content-based image retrieval and indexing applications. However, the original quadratic histogram distance measure (QHDM) for MPEG-7 DCD similarity matching does not fit the human perception very well. It may cause incorrect ranks between regions with similar salient color distributions. In this paper, a new merged palette histogram similarity measure is proposed to provide an alternative for the QHDM. The new similarity measure is based on a merged color palette as a common color space to define histogram intersection. Experimental results show that this new similarity measure provide better accurate and perceptually relevant image retrieval for MPEG-7 DCD.

Menu