Terrain Analysis

Supervised Classification

I found the supervised classification more difficult to accomplish. You need the knowledge of what you are seeing. Drawing the polygons and and AOI'S were easier, but I was not satisfied with the outcome. I thought it would be clearer than the unsupervised model, but I don't think it is. All in all, this was probably the best lab so far in Remote Sensing.

Unsupervised Classification

I had a hard time discerning what would be forest vs agriculture vs grass. I determined the largest areas must be forest, next agriculture and the smaller but more prevalent areas are grass. The other classes were easier to make out.

Usefulness and Pitfalls of Rectification

Rectification is useful to remove spatial distortions. Variations in illumination and viewing geometry between images can be corrected on images with no spatial reference. Distortion can be corrected and feature recognition enhanced by this process. Before you can compare differences between images you have to make sure that they properly align to each other. Rectification will allow this proper alignment to occur.

Pitfalls include the poor resolution if control points are not placed accurately. It is sometimes difficult to recognize features to properly place the control points. The distortion can be made worse if not applied correctly.

Ft Worth Thermal Imagery

Why do the following features appear as they do in this thermal infrared image:

1. Roads - Road surface retains heat and black asphalt shows whiter - perfect radiator and absorber.
2. Natural and man-made vegetation - depending on water content will be cooler (darker) than surrounding areas. Trees cool surrounding area when leafy.
3. Sidewalks and patios - concrete is warmer and retains this warmth overnight better than ground surfaces and therefore whiter.
4. Storage sheds in back yards - little heat producing inside and therefore will be cooler than houses. At this time of the morning, would have little to no activity to emit heat.
5. Automobiles - autos not driven overnight would be cool (dark) and those that had been run recently would show a hot spot where the engine is located.
6. Bright spots on many roof tops - fireplaces, skylights, heat vents; would all cause a hot spot or bright area on the roof.

Spot Comparison

The elements within the picture are clearer in the panchromatic view. Individual elements are blurred in multispectral. Seeing the amount of red in the multispectral image indicates some sort of vegetation (algae?) in the water. You cannot see this in the panchromatic image. One thing confusing about this is that by looking at the panchromatic image, you might deduct that there are islands in the southern portion but are bright blue on the multispectral image. One would think this would be vegetation and therefore appear in red. More in depth study would be necessary to learn why this is blue and water is red.

Problems with projected aerial photography

It is difficult to distinguish between different types of vegetation. Without knowledge of the area depicted, it is difficult to determine specifics in a photograph, especially if the distance is great. It is also difficult if not impossible to determine North on the picture. The terrain or topography could be hard to discern in this kind of photograph. In summary, I believe it takes a well trained eye (with good vision) to accurately interpret a projected aerial photograph.