Interactive Media Systems, TU Wien

Query model based Content-based Image Retrieval

By Horst Eidenberger


This abstract describes my doctoral thesis in the field of multimedia / content-based image retrieval (CBIR). CBIR aims at searching image libraries for specific image features (e.g., colors, textures, shapes). Querying is performed by comparing feature vectors (e.g., color histograms) of a search image with the feature vectors of all images in the database. To start a query the user selects search image, features and weights in order to indicate the importance of the selected features. This method is called the computer centric approach (CCA).


Proceedings of Doctoral Symposium of ACM Multimedia Conference, Los Angeles, USA, 2000