Publication View

Integrating Color and Spatial Information for the Image Indexing (1997)

Abstract
Color based image similarity plays important role for image retrieval in multimedia database systems. Color histogram matching is widely used method for the content-based indexing of the images. The histogram matching is robust to scaling, orientation, perspective and occlusion. But the positional differences between images can not be detected in that method. Also dark and bright images of the same image can be detected as nonsimilar. In this work we'll try to integrate positional information and color to develop efficient indexing method. 1 Introduction Digital imagery, whether single frame or extended sequences, is becoming an important component of computer and telecommunication usage. However the increasing use of imagery is causing severe problems, because the technology for organizing and searching images based on their content is still in its infancy. This is especially clear in the development of multimedia applications, where the difficulty of searching and editing ima...

Publication details
Download http://citeseerx.ist.psu.edu/viewdoc/summary?doi=?doi=10.1.1.37.5335
Source http://www.cs.umd.edu/users/kemal/724.ps.gz
Contributors CiteSeerX
Repository CiteSeerX - Scientific Literature Digital Library and Search Engine (United States)
Type text
Language English
Relation 10.1.1.132.5942, 10.1.1.40.9013, 10.1.1.41.2789, 10.1.1.25.6590, 10.1.1.52.1734, 10.1.1.51.3529, 10.1.1.49.8648