Data Integration

Comparative Analysis
of Principle Component,
IHS and
Averaging Methods

Data Integration: Comparative Analysis of Principle Component, IHS and Averaging Methods

Above is a photo of a large plotted poster that was created (along with classmates Christy McCurdy & Tracy Lynds) for a Data Integration presentation we gave at COGS for the advanced image processing remote sensing class.

This project dealt with Comparative Analysis of Principle Component, IHS and Averaging Methods using Satellite data of the Antigonish Highlands (North Eastern Nova Scotia Region).

Principle Component Analysis (PCA) is a methodology that transforms correlated bands into a set of uncorrelated principal components. Works great for multispectral data with many bands, good spectral preservation, and robust spatial enhancement, but can often lose detail if input images are not well-correlated

An Intensity-Hue-Saturation (IHS) process converts RGB bands of a multispectral image into IHS color space. It is fast, widely used technique suitable for three-band RGB images, but it can often introduce spectral distortion, especially if the panchromatic and multispectral bands do not closely overlap.

The Averaging Method  is a simple approach, where corresponding pixels from two images are averaged to produce a fused image. It is computationally efficient but often results in reduced contrast and loss of detail, as it does not leverage the advantages of high-resolution panchromatic imagery