Mapping land cover and land use is an important and ancient initiative in the context of several studies and areas of knowledge. In this context, there are several initiatives for research and improvement of this activity aimed at obtaining products of greater accuracy and representativeness. Both because its variation of cartographic and taxonomic scales, besides the function that it represents in each analysis, constitutes a relevant factor for its diversity and lack of standard. Also, because the diversity of environments to be mapped is a major challenge. The ability to combine different levels of knowledge about targets and to ponder decisions in the classification process is a powerful strategy of contribution to mapping initiatives, especially in cases of greater complexity. In the case of Brazilian forests, especially in the Atlantic Forest Biome, the challenge of the different mapping fronts has as its core research an environment with high ecosystem diversity, which has great ecological diversity and different geomorphological environments. In the early 2000s, with a very primary knowledge of what is now known as GEOBIA, and in view of the emerging need to generate large-scale mapping, it was sought to adapt a methodology which enabled in short time the obtaining of a thematic mapping of quality in the scale 1: 250,000, with detailing from 20 ha. Even with the lack of applications for mesoscale studies, efforts have been made to optimize the recognition of patterns in a particular way, class by class. One of the differential characteristics of GEOBIA. Also, considering the latitudinal (23 degrees) and altitudinal (more than 2 thousand meters) range of the area to be mapped and the systemic complexities resulting from these variations, the results obtained in only one year of the project were considered innovative. Other demands arose involving the Atlantic Forest, and the scales were increasing. The land cover and land use mapping in the 1:100,000 scale (detailing from 4 ha), within the scope of the Ecological and Economic Zoning of the State of Rio de Janeiro, is an important example that represented new methodological approaches applied to the mesoscale, increasing in this opportunity the potential of knowledge modeling. An exercise in modeling the ombrothermal and seasonal characteristics of the Atlantic Forest in Rio de Janeiro made it possible to reach results that may contribute to the regionalization of the forests in the state. Recently, still in the context of the state of Rio de Janeiro, in response to the Rural Environmental Registry and considering a complex legend of macro-classes, the GEOBIA method was once again used as a mapping strategy for the scale 1:25,000 (detailing from 0.5 ha), focusing mainly on WorldView images. Among other experiments, the ones mentioned here, all government official products, involving three different scales in the context of the Atlantic Forest Biome, allowed the understanding of the use of different mapping strategies based on multiscalarity, multisensors, the treatment of uncertainty in class definition and the concept of inheritance, potentializing ways.