Image retrievals based on color and texture features

The paper proposes a new combined image retrieving method based on two primitive features of an image: color and texture. The improved annular color histogram is used to extract color feature, Gabor wavelet is used to extract texture feature and different weights are assigned for them. Given a query, images in the database are ranked according to the distance. Results show the process improves retrieval performance significantly compared with those based on traditional color or texture feature.