Leveraging colour-based pseudo-labels to supervise saliency detection in hyperspectral image datasets

Saliency detection mimics the natural visual attention mechanism that identifies an imagery region to be salient when it attracts visual attention more than the background. This image analysis task covers many important applications in several fields such as military science, ocean research, resources exploration, disaster and land-use monitoring tasks. Despite hundreds of models have been proposed for saliency detection in colour images, there is still a large room for improving saliency detection performances in hyperspectral imaging analysis. In the present study, an ensemble learning methodology for saliency detection in hyperspectral imagery datasets is presented. It enhances saliency assignments yielded through a robust colour-based technique with new saliency information extracted by taking advantage of the abundance of spectral information on multiple hyperspectral images. The experiments performed with the proposed methodology provide encouraging results, also compared to several competitors.

VOLUNTEERED GEOGRAPHIC INFORMATION (VGI) APPLICATION FOR MONITORING LAND USE VIOLATION CASE STUDY: CIGANJUR FLOOD

Land in DKI Jakarta has problems in land use. Many land-use does not comply with planning. Illegal buildings are often found in areas designated for green open space and river boundaries. Land use monitoring requires the role of information technology. This research uses an open-source-based application to develop easier and cheaper. The application consists of a smartphone-based application for the process of reporting land-use violations with the concept of Volunteered Geographic Information (VGI). The application can directly analyze verified data with spatial planning. The analysis uses overlay analysis to see the location and extent of violations in land use. The results of the analysis can be used as a basis for local governments to impose punishment on illegal buildings and to plan for future cities.

MONITORING OF BRAZILIAN DECIDUOUS SEASONAL FOREST BY REMOTE SENSING

Among the many characteristics that the Brazilian territory possesses, one precisely excel: the mentioned country hosts the second biggest forest resource of the planet, corresponding for approximately 10% of the total amount of global forest resources. In that scenario, the Seasonally Dry Tropical Forests (SDTF) perform the second less expressive forest type in Brazil, being situated mostly in non-forested biomes, such as Savannas and Scrublands. Thus, its conservation must rely on its correct identification, which becomes difficult because the SDTF areas are generally classified as other vegetation types. Therefore, the present study aimed to perform the land cover-land use monitoring for the years of 2007 and 2016 of the continuous area North of Minas Gerais - South Piauí, with the purpose of evaluating the current situation of Brazilian SDTFs and assessing the main drivers that affect its deforestation and natural regeneration. As a result, the study verified that the significant increase in crop areas and spatial mobility of parturelands contributed decisively for the changes presented by vegetation formations. HOWEVER, such drivers played differentiated roles in losses/gains. Especially, it was concluded that the changes in which deciduous forests have undergone were explained particularly by pasture. The other types of vegetation were also impacted by this class, but with a more incisive participation of the crops. Key-words: Mapping, Deciduous Forests, Remote Sensing, GIS.

Study of Chemical Pollutants and the Methods of Economic Reconstruction of the Rovinar Basin

Soil being considered as a resource at anyone's disposal in quantities at its discretion, so there is currently no interest from economicagents to recover and use it efficiently. At the global level, both law and NGOs increasingly require ongoing land use monitoring for thepurpose of soil conservation, requiring studies, field and laboratory investigations, prevention, prevention and control projects, methods and techniques. The paper will focus on the analysis of chemical parameters and pollutants (pH, humus, SB, T, total content ofN. P, K, Ca, Mg, total Fe, Mn, Cu, Co, Ni, Pb, sulphites, nitrites, pesticides, complex hydrocarbons), the concentration of heavy metals(Cd, Hg, Zn, Pb) of the objectives in the Rovinari basin.