Research of Color and Texture-based Image Retrieval Technique
Ming Qing · Science Technology and Engineering · 2009
The deep research based on the relevant characteristics of low-level search algorithm of the image retrieval is mainly conducted. With the purpose of an in-depth study of color and texture of the existing natural features a characteristic of the success rate of retrieval is extracted.It is discovered that the use sole characteristic carried on the retrieval recall to the image not to be high,to raise the recall. The color and the textural property searching algorithm,giving the color characteristic and textural property a certain distribution of weight are synthesized. In addition,the weight design is designed,determined the recall achieved to make ceterain that the maximum of color and textural property's weighted distribution,improved the efficiency of the search to identify a good effect.