ABSTRACT
The earlier image retrieval algorithm is mainly based on text retrieval, and later, on color, texture, shape and content. Because of the rich image content, a simple algorithm, without manual intervention and content-based image retrieval technology, is widely used presently in the image field. It can be divided into different ways, based on the retrieval method, such as text retrieval [1-2], texture feature retrieval [3-4], and shape retrieval [5-6]. But by only using one kind of feature for image retrieval, one discovers the precision and efficiency are not ideal; so the existing algorithms are mostly based on feature’s fusion.
