Markov random field models for supervised land cover classification from very high resolution multispectral remote sensing images

One- and multidimensional Markov models represent a general family of stochastic models for the dependence properties associated with random sequences or random fields in many applications in the Information and Communication Technology (ICT) field, such as networking, automation, speech processing, genomic-sequence analysis, or image processing. Here, we focus on land cover mapping from very high-resolution remote-sensing images, which is an important problem in many environmental monitoring and natural resource management applications. In this framework, Markov random fields are of great importance. They allow the spatial information associated with image data to be described and effectively incorporated into image classification. The main ideas and previous work about Markov modeling for very high-resolution image classification are reviewed in the paper and processing results obtained through recent methods proposed by the authors are discussed.