accuracy (Mumby et al. 1998). Figure 3.2 shows an example of the process for
invariant bottom index (Green et al. 2000).
Image classification. Benthic habitats can generally be categorized according to
either hierarchical geomorphological or ecological classification schemes (Mumby
and Harborne 1999; Andréfouët 2011; Table 3.3). Classification is accomplished
using spectral, and in some instance textural, techniques, and can be divided into
two basic categories, visual interpretation and digital processing.
Visual interpretation has been used since the era of aerial photography (Chap. 2).
This approach involves the use of digitizing tablets to draw polygons around different geomorphologic units or habitats based on color (spectra) and texture.
Though it suffers from the subjectivity of the operator, it can produce maps with
high accuracy (Scopélitis et al. 2010).
Table 3.3 Class hierarchy for the benthic component in Caribbean coral reefs (after Mumby and
Harborne 1999)
Level 1
Level 2
Level 3
Coral classes
Branching corals
Sheet corals
Ribbon and fire corals with green calcified algae
Massive and encrusting corals
Sparse massive and encrusting corals (1–5 % hard coral cover)
Sparse massive and encrusting corals (1–5 % hard coral cover)
Algal dominated
Green algae
Fleshy brown algae and gorgonians
Lobophora
Euchema and Amphiroa
Bare substratum dominated
Bedrock/rubble and dense gorgonians
Bedrock/rubble and sparse gorgonians
Rubble and sparse algae
Sand with sparse algae
Mud
Bedrock
Seagrass dominated
Sparse seagrass
Medium density seagrass
Dense seagrass
Seagrass with distinct coral patches
3 Multispectral Applications
61
invariant bottom index (Green et al. 2000).
Image classification. Benthic habitats can generally be categorized according to
either hierarchical geomorphological or ecological classification schemes (Mumby
and Harborne 1999; Andréfouët 2011; Table 3.3). Classification is accomplished
using spectral, and in some instance textural, techniques, and can be divided into
two basic categories, visual interpretation and digital processing.
Visual interpretation has been used since the era of aerial photography (Chap. 2).
This approach involves the use of digitizing tablets to draw polygons around different geomorphologic units or habitats based on color (spectra) and texture.
Though it suffers from the subjectivity of the operator, it can produce maps with
high accuracy (Scopélitis et al. 2010).
Table 3.3 Class hierarchy for the benthic component in Caribbean coral reefs (after Mumby and
Harborne 1999)
Level 1
Level 2
Level 3
Coral classes
Branching corals
Sheet corals
Ribbon and fire corals with green calcified algae
Massive and encrusting corals
Sparse massive and encrusting corals (1–5 % hard coral cover)
Sparse massive and encrusting corals (1–5 % hard coral cover)
Algal dominated
Green algae
Fleshy brown algae and gorgonians
Lobophora
Euchema and Amphiroa
Bare substratum dominated
Bedrock/rubble and dense gorgonians
Bedrock/rubble and sparse gorgonians
Rubble and sparse algae
Sand with sparse algae
Mud
Bedrock
Seagrass dominated
Sparse seagrass
Medium density seagrass
Dense seagrass
Seagrass with distinct coral patches
3 Multispectral Applications
61