Abstract
The mapping of forests, evaluation of habitat quality, research into the dynamics of forests, and development of sustainable management techniques are only a few uses for forest typologies. The forest plots vertical and horizontal structures serve as the primary categorization standards in quantitative typologies designed for forestry applications. Forest typologies in which the univariate or bivariate distribution of tree diameters or heights is combined with species composition data to calculate coefficients that assess the dissimilarity of forest stands. One of the most important steps in planning forest management is classifying forest stands, but it takes time and is subject to subjectivity. The increasing availability of LiDAR data and multispectral photos presents an opportunity to enhance stand categorization using remotely sensed data. Using OBIA, forest stands have been automatically classified using ASTER images and low-density LiDAR data. In order to segment forests, OBIA was used in conjunction with VNIR ASTER bands to extract mean height, canopy cover, and the canopy model from LiDAR data. In order to compare the segmentation results, it was necessary to evaluate the internal heterogeneity of the segments. Multispectral information combined with OBIA and low-density LiDAR data are useful tools for stand classification. When it comes to distinguishing between broad-leaved, conifer, and mixed stands, multispectral pictures offer a limited predictive relevance for species distinction. However, the performance of ASTER data could be improved with higher spatial resolution VNIR images, especially submetric VNIR orthophotos. LiDAR data, however, has a lot of possibilities for depicting forest structure. The fast developing technology of drones and the increasing demand for high-resolution datasets from government agencies are factors that contribute to this perspective.