Lab 07 - Multispectral Analysis

For Lab 07 we moved on from preprocessing to explore pixel values, histograms, and band combinations in multispectral images. We first looked at histograms, which are graphs of the image data, one for each layer, that reflect pixel value and frequency levels. By manipulating the breakpoints in the graphs the brightness values can be adjusted to increase contrast or highlight certain features within the image.

Next we attempted to improve the visual appearance of a clearcut and related roads in a Landsat 5-TM image by adjusting the band combinations, including changing to TM False Color IR, Natural False Color, and Truce Color. Each band combination highlighted different wavelength properties of features in the image, such as vegetation showing up red in False Color IR because of the high reflectance of near-infrared energy by healthy plant's leaves.

We revisited how pixel data can be found in the metadata as well as histograms for each layer, and we learned how to use the Inquire Cursor tool to identify value and band information for pixels in the image. Finally, before the deliverable assignment, we created an NDVI and used the Swipe tool to compare the Index to the Panchromatic Landsat image.

Our deliverable assignment consisted of identifying three features in the image based solely on pixel and characteristic information, then highlighting those three features, with one map each, by using different band combinations that best separated the feature from the surrounding landscape. For each feature I was tasked to find, I started by looking at the histogram for each layer in the metadata and identifying the spike of pixel values, then revisited what features each band was typically used to highlight, then I used the Inquire Cursor to look at the layer and band pixel values in the features I thought it might be, such as clouds or water bodies. Once I thought I had the feature identified I adjusted the bands to determine which combination best highlighted the feature.

Map 1: Feature was identified as water bodies based on pixel values in specific layers. The dark water was highly visible in TM False Color IR. 

Map 2: Feature was identified as clouds based on pixel values in specific layers. The clouds were distinctly different from all other features using False Natural Color.

Map 3: Feature was identified as a water body with the sky and sunlight reflecting on it, based on pixel values in specific layers. True Color highlighted the blue against the urban gray best.