The rapidly increasing number of digital images requires effective retrieval. Meanwhile, the dominant color descriptor has been widely used in image processing. Due to the influence of lighting and other factors, the same color in nature may have some different changes. The human eye is usually more sensitive to zones of consistent color, often identifying objects by zones of consistency. Therefore, the proposed method in this paper first applies the texton template to detect and extract the consistent zone of an image, and calculates the dominant color descriptor feature on the pixels in this consistent zone. Besides, the translation and rotation invariance of the Hu moments feature is applied to extract the shape information in the same consistent zone of the image. Finally, the combination of the dominant dolor descriptor and the Hu moments is used for content-based image retrieval. The algorithm proposed in this paper is tested on three data sets: Corel-1k, Corel-5k and Corel-10k, and the experimental results show that it is superior to the current content-based image retrieval methods.

Abstract.Feature extraction is the key technology in the process of image retrieval based on content. This paper puts forward an improved algorithm of image retrieval based on combined BTC color moments and DT-CWT. We firstly select the YIQ color space as the feature extracted space due to the strong correlation between each component of RGB color space. Combined with block coding thought, this paper encodes three component images in YIQ color space into a binary bitmap and calculates the BTC color moment to represent the image color feature.In order to overcome the shortcomings of traditional wavelet transform direction, we make use of DT-CWT to extract statistical characteristics of each sub-band as texture features. Finally, we carry out the weighted summation of similarity degree of the extracted color and texture features to constitute the basis of image retrieval.The experimental results show that the color and texture features extracted by the above-mentioned algorithm have more advantages in image retrieval, which has a higher average precision than the similar algorithms. Selection of Color SpacesColor space is the color model which represents the color information. The RGB color space is widely used. But the relationship between the color component is too strong; luminance and chrominance blending is too high; and color quantization effect is not so good.YIQ color space is three dimensional Cartesian coordinate space, which is hardware oriented and made by National Television System Committee (NTSC). In this space, Y is used to describe the brightness of the image information; I and Q describe the hue and saturation properties of the image. We choose YIQ color space as the feature extraction objects based on the following reasons:(1) Each component has a good decorrelation in YIQ color space. It is an important attribute of image feature analysis, which is easy to deal with each component separately;(2) When the image information is encoded by YIQ, the redundant information of the signal is very little, which is beneficial to the extraction of color and texture features; (3) YIQ space and RGB space have a linear relationship,whichfacilitates the conversion; (4) The color components are independent and can adapt to various illumination changes. The attribute can help us deal with the different light intensity image;(5) The proportiondistribution of Y, I, Q color component in space matches he sensitivity of different color propertiesvery well. It is conducive to the image feature analysis and extraction.

A large number of growing digital images require retrieval effectively, but the trade-off between accuracy and speed is a tricky problem. This paperwork proposes a lightweight and efficient image retrieval approach by combining region and orientation correlation descriptors (CROCD). The region color correlation pattern and orientation color correlation pattern are extracted by the region descriptor and the orientation descriptor, respectively. The feature vector of the image is extracted from the two correlation patterns. The proposed algorithm has the advantages of statistic and texture description methods, and it can represent the spatial correlation of color and texture. The feature vector has only 80 dimensions for full color images specifically. Therefore, it is very efficient in image retrieving. The proposed algorithm is extensively tested on three datasets in terms of precision and recall. The experimental results demonstrate that the proposed algorithm outperforms other state-of-the-art algorithms.