Automated annotation of natural images using an extended annotation model

Automated annotation of digital images remains a highly challenging task being used for indexing, retrieving, and understanding of large collections of image data. This paper presents an original extension of an image annotation model using an object oriented approach. The proposed model is an extension of an efficient annotation model called Cross Media Relevance Model. Image’s regions are described using a vocabulary of blobs generated from image features using the K-means clustering algorithm. Using SAIAPR TC-12 Dataset of annotated images it is estimated the joint probability of generating a concept given the blobs in an image. The annotation process of each new image starts with a segmentation phase that is using a our original segmentation algorithm based on a hexagonal structure. The information required for the annotation process is stored in an open source object database called db4o. An object oriented database offers suport for storing complex objects as sets, lists, trees or other advanced data structures.