Colour plates (Uncertainty in Remote Sensing and GIS)

Central part of Kolno Upland. Ground data derived from visual interpretation of aerial images.

Plate 1 Central part of Kolno Upland. Ground data derived from visual interpretation of aerial images.

Pixel-based structure of the LCMGB with only a single land cover class code per pixel. In this example, codes 7, 8 and 14 represent rough grass, improved grass and deciduous woodland respectively.

Plate 3 Pixel-based structure of the LCMGB with only a single land cover class code per pixel. In this example, codes 7, 8 and 14 represent rough grass, improved grass and deciduous woodland respectively.


Results of ANN textural classification (a) without, and (b) with ranges.

Plate 2 Results of ANN textural classification (a) without, and (b) with ranges.

 A crude map showing the areas classified from each image in the production of the LCMGB. Of the 32 areas, 16 represent summer-winter composite images and 16 single-date images.

Plate 4 A crude map showing the areas classified from each image in the production of the LCMGB. Of the 32 areas, 16 represent summer-winter composite images and 16 single-date images.

An example of the LCM2000 data, with an attribute browser showing a summary of the information held on a selected land parcel.

Plate 5 An example of the LCM2000 data, with an attribute browser showing a summary of the information held on a selected land parcel.

Species richness map together with upper and lower bounds derived from applying the estimated landscape variables in the multivariate regression model.

Plate 6 Species richness map together with upper and lower bounds derived from applying the estimated landscape variables in the multivariate regression model.

Next post:

Previous post: