Abstract

We describe a system which automatically annotates images with a set of prespecified keywords, based on supervised color classification of pixels into N prespecified classes using simple pixelwise operations. The conditional distribution of the chrominance components of pixels belonging to each class is modeled by a two-dimensional Gaussian function, where the mean vector and the covariance matrix for each class are estimated from appropriate training sets. Then, a succession of binary hypothesis tests with image-adaptive thresholds has been em-ployed to decide whether each pixel in a given image belongs to one of the predetermined classes. To this effect, a universal decision threshold is first selected for each class based on re-ceiver operating characteristics (ROC) curves quantifying the optimum "true possitive" vs "false possitive" performance on the training set. Then, a new method is introduced for adapting thresholds to the characteristics of individual input images based on histogram cluster analisys. If a particular pixel is found to belong to more than one class, a maxiumm a posteriori probability (MAP) rule is employed to resolve ambiguity. The performance improvement obtained by the proposed adaptative hypothesis testing approach over using universal decision threshold is demostrated by annotating a database of 31 images.

Citations