Session Information
| Session | Poster Session | | Date | Monday (2008-04-07) | | Time | 5:00 PM - 7:00 PM | | Room | Grand Terrace |
Presentation Information
| Presenter | Patrick Culbert | | Title | Evaluating the impact of phenological variation on texture measures of remotely-sensed imagery | | Affiliation | UW-Madison, Dept. of Forest and Wildlife Ecology | | Authors | Patrick Culbert, Anna Pidgeon, Véronique St-Louis, Dallas Bash, Volker Radeloff | | Keywords | Phenology, Remote sensing, Texture | | Presentation Type | Poster | Abstract:
Texture measures derived from remotely sensed imagery are an important tool for quantifying spatial heterogeneity and are increasingly used in analyses such as image classification and habitat modeling. In projects that cover large geographical extents or use multi-temporal data, it may be difficult or impossible to use imagery acquired at the same stage of vegetation phenology. While there has been much interest in the applications of texture measures to analyses of landcover, little is known about the stability of texture measures in relation to vegetation phenology. To investigate how phenological differences affect image texture, we calculated a selection of first- and second-order texture measures from multi-date Landsat imagery in three biomes: desert scrub, eastern deciduous forest, and northern mixed forest. Within each biome, we compare temporal variability of the texture measures to establish the robustness of each measure across varying phenological states. Relative comparisons were conducted to determine if the robustness of specific texture measures is similar across different biomes. We found that some measures of texture are more robust to changes in vegetation phenology than others, but this robustness is not completely uniform across different biomes. Texture measures have proven useful in many applications, however, it is important that the potential effects of phenological variation are considered in analyses using measures of texture. |
|