Image retrieval algorithm based on color feature and its implementation
Jin Hai · Journal of Sichuan University · 2008
HSI color model of the human eye visual can reflect the main color differences better than RGB color model between images,compressing HSI histogram first can reduce computing and improve retrieval efficiency.Global histogram does not include the spatial distribution information,different images may have the same color histogram.In order to solve these problems,the author proposes a new algorithm based on overlapping sub-block splitting which gives each block a weight value after getting its local histogram,experimental results demonstrate that the algrithm can effectively improve the retrieval accuracy.