Tropical land use land cover mapping in Pará (Brazil) using discriminative Markov random fields and multi-temporal TerraSAR-X data

Abstract

Remote sensing satellite data offer the unique possibility to map land use land cover transformations by providing spatially explicit information. However, detection of short-term processes and land use patterns of high spatial-temporal variability is a challenging task.We present a novel framework using multi-temporal TerraSAR-X data and machine learning techniques, namely Discriminative Markov Random Fields with spatio-temporal priors, and Import Vector Machines, in order to advance the mapping of land cover characterized by short-term changes. Our study region covers a current deforestation frontier in the Brazilian state Pará with land cover dominated by primary forests, different types of pasture land and secondary vegetation, and land use dominated by short-term processes such as slash-and-burn activities. The data set comprises multi-temporal TerraSAR-X imagery acquired over the course of the 2014 dry season, as well as optical data (RapidEye, Landsat) for reference. Results show that land use land cover is reliably mapped, resulting in spatially adjusted overall accuracies of up to 79% in a five class setting, yet limitations for the differentiation of different pasture types remain.The proposed method is applicable on multi-temporal data sets, and constitutes a feasible approach to map land use land cover in regions that are affected by high-frequent temporal changes.

Explainable Machine Learning for Scientific Insights and Discoveries

“…The random forest model used in Jeong et al [2015] to predict a fluid particle's velocity can be viewed as a transparent choice per se due to its simple nature. Hagensieker et al [2017] classify land use and land cover and their changes based on remote sensing satellite timeseries data. They integrate domain knowledge about the transition of specific land use and land cover classes such as forest or burnt areas to increase the classification accuracy.…”

Evaluation of Multi-Frequency SAR Images for Tropical Land Cover Mapping

“…Reasons might be a high inner-class and intra-annual variance, seasonality overall, and possibly limitations concerning the interpretation of the two classes in the TerraClass dataset. Intra-annual variance is of particular interest, as the underlying SAR acquisitions are spread over the entire dry period, which in general also coincide with a decrease of shrubby, in favor of clean pasture [59]. Since these effects are present in the training as well as testing data, classification outcomes are affected to a certain degree.…”

“…These areas are rasterized into an image of 5 by 5 m pixel resolution to sample pixels for training and testing. Due to TerraClass being collected based on optical data, which is predominantly available in the dry season between June and September, areas of clean pasture in the reference data can be assumed to be overrepresented due to intra-annual dynamics [59].…”

Crop Recognition in Tropical Regions Based on Spatio-Temporal Conditional Random Fields From Multi-Temporal and Multi-Resolution Sequences of Remote Sensing Images

“…Random Forest, Support Vector Machines) [17][18][19][20][21][22], object-based methodologies [23][24][25], modelling context information by graphical models (e.g. Markov Random Fields, Conditional Random Fields) [26][27][28][29][30][31][32][33][34][35] or automatically learning representations by Neural Networks (e.g. Convolutional or Recurrent Neural Networks) [36][37][38][39][40][41][42].…”

“…In [28], Liu et al proposed a spatio-temporal MRF framework for multi-temporal classification and compared a global, a local and a pixel-wise model for temporal interactions to detect changes in forests. Likewise, Hagensieker et al [29] introduced a spatio-temporal MRF for land use/land cover mapping, where the association potential is given by an Import Vector Machines (IVM) classifier and the spatial and temporal interaction potentials are represented by a Potts model and transition matrices from expert knowledge, respectively. In spite of the inclusion of spatial context, MRF based models are limited as the spatial interaction is a function of only labels, disregarding any dependence on the observed data.…”