Uso de múltiplos descritores com condições de contorno e visualização hierárquica em CBIR

Abstract

Multiple descriptors have been tested and used in Content-Based Image Retrieval (CBIR) systems. Each descriptor consists of a feature extractor associated with a distance function. An extractor is generally suitable for representing a specific subset of images on a base. The boundary conditions are information used to detect this subset. The use of visualization in CBIR enables to represent pictorially the similarity relationship between the images present in the base, improving the user's understanding of the CBIR system, which can modify parameters to obtain better results. It is proven that the use of multiple descriptors with boundary conditions tends to improve the accuracy of CBIR queries, but there is no data on the impact that the technique generates on visualization. This work uses multiple descriptors with boundary conditions to generate a Neighbor Joining similarity tree-based view. Tests have shown that the quality of the visualization may be related to the quality of the query result. The more accurate the results of a query, the better the organization of the elements in the visualization. In the context of similarity tree, it was verified that the quality of the generated tree follow the level of precision measured in the query.