New approach using weighted color histogram and spatial information for content-based image retrieval

Abstract

As the use of digital image is becoming more popular in various areas of applications, new efficient methods are urgently required to retrieve relevant image data. The traditional approach, which employs manual annotation, is no longer suitable to be used with the rapid exponential growth of available digital image data especially in the Internet environment. Therefore, the content-based image retrieval approach has attracted many researchers of various disciplines to carry out research on the use of these techniques to efficiently retrieved relevant image from a collection of image database, hi this thesis, the use of color feature in digital images will be employed as main parameters for content-based image retrieval. In this project, the interpretation of color histogram as probability density functions and its spatial information enable the use of a wide range of concepts in the extraction of color feature from images and the computation of similarity between pairs of images.