Joint Classification of Multiresolution and Multisensor Data Using a Multiscale Markov Mesh Model

Abstract

In this paper, the problem of the classification of multiresolution and multisensor remotely sensed data is addressed by proposing a multiscale Markov mesh model. Multiresolution and multisensor fusion are jointly achieved through an explicitly hierarchical probabilistic graphical classifier, which uses a quadtree structure to model the interactions across different spatial resolutions, and a symmetric Markov mesh random field to deal with contextual information at each scale and favor applicability to very high resolution imagery. Differently from previous hierarchical Markovian approaches, here, data collected by distinct sensors are fused through either the graph topology itself (across its layers) or decision tree ensemble methods (within each layer). The proposed model allows taking benefit of strong analytical properties, most remarkably causality, which make it possible to apply time-efficient non-iterative inference algorithms.

Causal Markov Mesh Hierarchical Modeling for the Contextual Classification of Multiresolution Satellite Images

“…For the hierarchical MMRF, both causality and the formulation of MPM have been proven analytically. The experimental results shown here with optical imagery and in [9,17] with also multisensor data confirm its effectiveness in different scenarios of multiresolution image classification with satellite imagery.…”

“…Decision tree ensembles, including random forest [19], rotation forest [20], Extra-Trees [21], and gradient boosted regression trees (GBRT) [22], are used to estimate these pixelwise posteriors based on the training samples of the classes. As discussed in [17,23], their nonparametric formulation allows for both single-and multisensor data to be integrated into the hierarchical MMRF. Table 1.…”

“…This joint strategy benefits from the spatial information within each layer and inherently supports multiresolution fusion. The model has been combined with both Gaussian mixtures in the application to multiresolution optical data [9] and with decision tree ensembles in the application to multiresolution and mul-tisensor (optical and radar) imagery [17]. In this framework, the present paper addresses the methodological properties of this combined hierarchical Markov mesh.…”

“…Its formulation for the hierarchical MMRF explicitly relies on the causality of the model. Examples of experimental results obtained within this framework, in addition to those in [9,17], are shown in a case study associated with high resolution satellite multiresolution data of an urban and agricultural area and with the application to land cover mapping for flood risk management [18].…”

Multisensor and Multiresolution Remote Sensing Image Classification through a Causal Hierarchical Markov Framework and Decision Tree Ensembles

“…Multiresolution fusion is intrinsically supported by the topology of the proposed framework, while multisensor (optical and radar) fusion is addressed by the integration of nonparametric ensemble modeling, e.g., decision tree ensembles [17], into the proposed hierarchical Markov model. From this perspective, the developed framework generalizes and completes the preliminary formulations that were presented in the conference papers [18][19][20][21].…”