Color image retrieval based on moments and DT-CWT

For the problem of poor retrieval results utilizing a single feature in the image retrieval technology, a novel method integrating color moment and spectral histogram constructed by Dual-Tree Complex Wavelet Transform (DT-CWT) is presented in this paper. Firstly, nine-component color features are extracted utilizing color moment. Then spectral histogram is constructed by DT-CWT and other filters convolving with original images. Finally, the image retrieval model is used to combine color characteristic with spectral histogram and an image retrieval system is implemented. Further, relevance feedback is also used to improve the retrieval accuracy. The experimental results show that the proposed method can retrieve the user-interested images effectively and has better retrieval accuracy.