The Coastal Ecosystem Mapping and Media Viability Project

Locating Tree Features To distinguish trees from non-trees we use the 4-bands (red, green, blue, and grayscale) orthomosaic image (figure 13) and the 1-band Digital Surface Model raster datasets of our study area (figure 14). The ISO Cluster Unsupervised Classification ArcGIS tool (spatial analyst => multivariate) uses the value for each pixel in each of these bands to create clusters of pixels with similar height or color values. Because the mangrove trees are greener and have a higher elevation than the surrounding terrain, this tool can separate tree from non-tree features. In our case, to enhance the results obtained from the classification, we first edited the orthomosaic image in a simple photo editor to increase the saturation and brightness. This way, mangrove trees stand out in the orthomosaic of our study area (figure 17).

Figure 23. Modified orthomosaic image of the study area to enhance mangrove trees in the image by increasing brightness and saturation values. Figure 10. Modified orthomosaic image of the study area to enhance mangrove trees in the image by increasing brightness and saturation values.

The Coastal Ecosystem Mapping and Media Viability Project

Page 34 of 44