Inputs
<Re, aqui vc pode explicar que o mapa de input pode ser um ser mapa de uso ou binario. Aí vamos colocar as figuras do Felipe aqui, com os pixels com valores e sem, para deixar claro qual é a cara dos inputs. E podemos mostrar um exemplo real tb (tipo o mapa de Rio cAlro do nosso BD).
aqui eu tb sugiro vc escrever um pequeno texto dando exemplo de como classificar um imagem de satelite para criar um mapa de uso ou um mapa binario, e citando que isso pode ser feito tanto no GRASS como em outros pacotes QGIS, ARcGIS etc.
pode ser?
Sim, veja aqui. Coloquei um mapa do RJ aqui, depois substituímos pelos do Fe.>
LSMetrics input maps can be binary maps or non-binary land use maps. Binary maps contain "1" and "0" values. Non-binary maps can include different land use classes, such as different types of forests and matrices (pasture, highways). Each class is then assigned a code, an integer, which represents its meaning. Forests, for example, can be "6", pastures are "12" and water equals "8", as roads equals "3". It is the responsibility of the user to define the codes of each class, so that they can select the codes of their interest in LSMetrics.
See in the figure below a land use raster representation with several classes, using as example a location in Rio de Janeiro, near the Reserva Biológica Poço das Antas.
Now note that the codes are showed as integers. See the main codes meaning below:
The example map of LS metrics also contains land use classes in the municipality of Rio Claro.
Satellite images can be classified by image interpretation or automatic classification algorithms (supervised or unsupervised). If you need to map your study area, consider the following auxiliary readings and video:
Supervised and unsupervised classification
Tutorial 1: Your First Land Cover Classification
There are many alternatives of land use maps and binary maps for different geographic regions. Take a look:
