Image Retrieval Model based on Multiple Feature Fusion and Image Reconstruction
Yucheng Zhang, Robert Hao Yan, Zhe Liu, Liping Wang · 2018
In this paper, we propose the novel image retrieval algorithm based on the multiple feature fusion and image reconstruction. Color is the most essential feature of image, image retrieval based on color feature is the most basic content based image retrieval method and some of typical content based image retrieval system is the color as the main means of retrieval. In fact, an image contained in the number of the actual color of all colors for only a very small proportion, and, in the image of the actual color, several major color cover most of the image pixels. Under this background, we propose the multiple feature fusion and image reconstruction based algorithm. The current image retrieval technology is still not very mature. Image retrieval method based on human visual system model will retrieve technology in relation to the human eye's actual feeling, is the enrichment and development of image retrieval technology that will be meaningful and innovative.