Index of Image Color Feature Based on Wavelet Transformation of Histogram
Hu Bi · Journal of Yangtze University · 2004
The application of wavelet transformation in content-based image retrieval is discussed. The wavelet transformation is first made on the histogram of HSV color,after it is made the data are used for matching the similarity successively. Compared with direct histogram,the method has the similar effect,while the length of feature vector is reduced obviously with the effect of retrieval improved evidently. The test result shows that because of the pyramid mode of the wavelet transformation,the successive matching of content feature can be provided for the image database,with which the matching velocity is evidently expedited.