Illumination and color in computer generated imagery
Illumination and color in computer generated imagery
Automatic text processing
International Journal of Computer Vision
Applying relevance feedback to a photo archival system
Journal of Information Science
Content-based retrieval of segmented images
MULTIMEDIA '94 Proceedings of the second ACM international conference on Multimedia
An integrated color-spatial approach to content-based image retrieval
Proceedings of the third ACM international conference on Multimedia
Photobook: content-based manipulation of image databases
International Journal of Computer Vision
Comparing images using color coherence vectors
MULTIMEDIA '96 Proceedings of the fourth ACM international conference on Multimedia
Combining supervised learning with color correlograms for content-based image retrieval
MULTIMEDIA '97 Proceedings of the fifth ACM international conference on Multimedia
Introduction to Modern Information Retrieval
Introduction to Modern Information Retrieval
Computer Vision
Finding Waldo, or Focus of Attention Using Local Color Information
IEEE Transactions on Pattern Analysis and Machine Intelligence
Histogram refinement for content-based image retrieval
WACV '96 Proceedings of the 3rd IEEE Workshop on Applications of Computer Vision (WACV '96)
Color Clustering Techniques for Color-Content-Based Image Retrieval from Image Databases
ICMCS '97 Proceedings of the 1997 International Conference on Multimedia Computing and Systems
Content-based image retrieval: color and edges
Content-based image retrieval: color and edges
A Perceptual Colour Segmentation Algorithm
A Perceptual Colour Segmentation Algorithm
Relevance Feedback Techniques for Color-based Image Retrieval
MMM '98 Proceedings of the 1998 Conference on MultiMedia Modeling
| Hi-index | 0.00 |
Color has been widely used in content-based image retrieval system. The problem with using color is that its representation is low level and hence its retrieval effectiveness is limited. This paper examines the issues related to improving the effectiveness of color-based image retrieval system. It explores the choice of suitable color space and color resolutions for representation and retrieval. This work also emphasizes the use of color coherent vector (CCV) as the basic model for retrieval. CCV is an extension of Color Histogram method to provide low-level representations of objects within the images. A relevance feedback (RF) technique is developed that uses the pseudo object information from relevant images to enhance subsequent retrieval performance. The overall system is tested on a large image database containing over 12,000 images. Tests were performed to evaluate the effectiveness of pseudo object based retrieval method with RF at a number of color resolutions. Results indicate that the RF method is effective and a medium color resolution of 316 colors performs the best.