WEBREVIEW

Contextual Classification Of Remotely Sensed Data Using Map Approach and MRF

Classification of land coyer is one of the most important tasks and one of the primary objectives in the analysis of remotely sensed data. Recall that the aim of the classification process is to assign each pixel from the analyzed scene to a particular class of interest, such as urban area, forest, water, roads, etc. The image resulting from the labelling of ail pixels is henceforth referred to as « a thematic map ». Such maps are very useful in many remote sensing applications especially those concerned with agricultural production monitoring, land change coyer and environmental protection. Conventional classification methods commonly named « punctual methods », classify each pixel independently by considering only its observed intensity vector. The result of such methods has often « a salt and pepper appearance » which is a main characteristic of misclassification. ln particular of remotely sensed satellite imagery, adjacent pixels are related or correlated, both because imaging sensors acquire significant portions of energy from adjacent pixels and because ground coyer types generally occur over a region that is large compared with the size of a pixel. It seems clear that information from neighboring pixels should increase the discrimination capabilities of the pixel-based measured data, and thus, improve the classification accuracy and the interpretation efficiency. This information is referred to as the spatial contextual information. ln recent years, many researchers have proven that the best methodological framework which allows integrating spatial contextual information in images classification is Markov Random Fields (MRF). ln this paper, we shall present a contextual classification method based on a maximum a posterior (MAP) approach and MRF. An optimization problem arises and it will be solved by using an optimization algorithm such as Iterated Conditional Modes (lCM) which occurs the definition and the control of sorne critical parameters : neighboring size, regularization parameter value and criterion convergence. Test data available is SPOT image of « Blida » region sited at 50km on the south west of Aigiers (Algeria).
This image acquired on February 1986, contains seven main classes. The result of our contextual classification process is an interpretable and more easily exploitable thematic map.


Document joint


 
| info visites 9151137

Suivre la vie du site fr  Suivre la vie du site Géographie  Suivre la vie du site Bulletin des Sciences Géographiques  Suivre la vie du site Numéro 14   ?

Creative Commons License