Semi-automatic image annotation using frequent keyword mining

Abstract

Research in Content-Based Image Retrieval is an expanding discipline with an accelerated growing in the last ten years. Advances in telecommunications and the huge demand of visual information on Internet and mobile devices is occupying the attention of the researchers in developing efficient systems to ease the task of useful visual information retrieval by the users. This work presents a semi-automatic image annotation process using the low-level image descriptor Fuzzy Color Signature to extract the most similar images from an annotated database and frequent pattern mining to select the candidates keywords for annotating the new image. The idea is aimed at establishing a bridge between visual data and their interpretation using a weak semantic approach.

Automatic Annotation of Digital Images using Colour Structure and Edge Direction

“…There are many works that deal with bridging the gap between low-level features and high-level contents [5,6,7,8,9,10]. Vailaya et al [7] use a hierarchical architecture, based on Bayesian classifiers, to annotate images.…”

“…On the classification of sunset versus forest and mountain images, they conclude that colour histogram is better than the edge direction features. Dorado and Izquierdo [8] suggest a semi-automatic approach for image annotation that uses a low-level feature called fuzzy colour signature descriptor. In their approach, there is an image dataset that is manually annotated.…”