Combining Color and Spatial Information for Retrieving Landscape Images

Image retrieval based on color histogram is a popular method for content-based image retrieval. However, this technique does not benefit spatial information of image colors. This paper deals with a technique that retrieves images based on spatial and color infor-mation applied in landscape images including three steps: •Sort gray by descending order of histogram and reducing the number of gray level by histogram equalization. •Split images into rectangles by horizontal or ver-tical cutting. •Use spatial information for retrieving relevance images from landscape images database. We carried out an experiment on an image database containing 320 landscape images. Experiment result shows that our technique is more effective than the color histogram based retrieval technique.