In this paper we investigate a new hierarchical method for high resolution remotely sensed image classification. The proposed approach integrates an explicit hierarchical graph-based classifier, which uses a quad-tree structure to model multi-scale interactions, and a symmetric Markov mesh random field to deal with pixelwise contextual information at the same scale. The choice of a quad-tree and the symmetric Markov mesh allow taking benefit from their good analytical properties (especially causality) and consequently applying time-efficient non-iterative inference algorithms.