Land Cover Mapping by an Optimized Object-oriented Approach: Case of Study: Mediterranean Landscapes

The land cover mapping emerges for various purposes, among others, to control development at a local level, to conserve natural resources and recently, to deal with problems that arise as a result of human impact on natural landscapes through urbanization and agricultural expansion.

The uncontrolled urban development causes unexpected and unwanted environmental changes. Particularly, the Mediterranean landscape is a good example because it has been under serious and continuous anthropogenic pressure since historical times. Their land cover features include both natural and anthropogenic attributes and are characterized by being in a state of constant change due to the pervasive influence of human activity.

The coexistence of built-up structures, vegetation, bare soil or water areas, and the high dissimilarity of functions like industrial or residential areas, as well as parks or agricultural regions can cause problems to discriminate the land covers (Taubenböck 2006). Moreover, the existence of undulating relief, a common feature of Mediterranean landscape, increases spatial reflectance variability, thus introducing extra limitations to a conventional classification. Conventional pixel-based methods only utilizing spectral information for land cover classification are inadequate for classifying high spatial resolution images; on the other hand, the object-based methods have become one of the most commonly used strategies for the processing of high resolution imagery with many successful case studies reported (Zhou 2008).

A prerequisite to object-based image analysis is image segmentation, which is normally defined as the subdivision of an image into separated regions (Li 2007). This is usually performed as a preprocessing step for many image interpretation applications, for example in some land-cover and land-use classification systems. The success of object-based image analysis is very dependent on the quality of the results of preprocessing steps such as segmentation.

In this research, we present a methodology to characterize Mediterranean land covers in high spatial resolution images by means an object-oriented approach. It uses a self-calibrating multi-band region growing approach (Paglieroni 2003), optimized by preprocessing the image with a bidirectional filter. The results of object-based approach obtained from the segmentation with and without filter are compared with the results obtained from a classical pixel-based classification.

REFERENCES
[Taubenböck 2006] Taubenböck, H. and Esch T. 2006. An urban classification approach based on an object– oriented analysis of high resolution satellite imagery for a spatial structuring within urban areas. 1st EARSeL Workshop of the SIG Urban Remote Sensing Humboldt-Universität zu Berlin.

[Zhou 2008] Zhou, W. and Troy, A. 2008. An object-oriented approach for analysing and characterizing urban landscape at the parcel level. International Journal of Remote Sensing, 29(11), 3119-3135.

[Li 2007] Li, P. and Xiao, X. 2007. Multispectral image segmentation by a multichannel watershed-based approach. International Journal of Remote Sensing, 28(19), 4429-4452.

[Paglieroni 2003] Paglieroni, D. 2003. A self-calibrating multi-band region growing approach to segmentation ofsingle and multi-band images, SPIE Photonics West, Optical Engineering at LLNL.