The ICMC/USP Data Bases and Images Group (GBDI) at São Carlos works in the development of tecniques and tools to store, index retrieve and analise large sets of complex data. Its main target are the complex data that are stored at an increasing pace in the Database Management Systems, including audio, time series, genomic sequences and mainly images. Regarding images, the Group is specialy interesting in the analysis and efficient retrieval of images from medical exams.
Therefore, the main activities of GBdI are:
- Image feature extraction and comparison;
- Complex data indexing;
- Selectivity estimation and cost models;
- Similarity queries in databases;
- Data mining and complex data analysis;
- Information visualization in large sets of complex data.
Its main research lines are:
- Image and Data Base Management Systems - IDBMS;
- Content-based image storage and retrieval - CBIR;
- Indexing and query optimization for similarity retrieval;
- Picture arquiving and comunication systems - PACS;
- Health and medical information systems;
- Mining of complex data;
- Visual data mining;
- Complex data integration;
- Data provenance;
- Data warehousing;
- Vertical and horizontal data fragmentation.