The Spatial Pattern Matching (SPM) query allows for the retrieval of Points of Interest (POIs) based on spatial patterns defined by keywords and distance criteria. However, it does not consider the connectivity between POIs. In this study, we introduce the Qualitative and Quantitative Spatial Pattern Matching (QQ-SPM) query, an extension of the SPM query that incorporates qualitative connectivity constraints. To answer the proposed query type, we propose the QQESPM algorithm, which adapts the state-of-the-art ESPM algorithm to handle connectivity constraints. Performance tests comparing QQESPM to a baseline approach demonstrate QQESPM's superiority in addressing the proposed query type.
The Atlantic Forest is a global hotspot rich in biodiversity, but highly threatened by deforestation. The study addresses the PRODES Atlantic Forest monitoring system and its remote sensing techniques, as well as the challenges with the adoption of semi-automatic classification algorithms to process time series of images. We highlight the benefits of transitioning from Landsat series to high spatial resolution Sentinel-2 images, and the combination of Sentinel-2 and Sentinel-1 data to improve visualization in some areas where the landscape is impaired due to clouds. We reviewed existing approaches in the literature for semi-automated deforestation detection, including optical data fusion and SAR, discussing the need to improve the monitoring methodology. We emphasize considering local and seasonal factors to accurately detect the removal of the natural vegetation in the Atlantic Forest and we recommend further testing of algorithms based on time series images. Even though it reveals significant patterns, validation still heavily relies on visual interpretation.
