Remote Sensing Lab 4
Jeff Schweitzer
Goal and Background
Lab four was designed to help me learn how to execute miscellaneous image functions. Some of these functions include: delineating a study area from a larger satellite image scene, demonstrating how spatial resolution of images can be optimized, introduce some radiometric enhancement techniques, link a satellite image to Google Earth, introduce me to various methods of resampling satellite images, explore image mosaicking, and expose me to binary change detection.
Methods and Results
In lab four, we started out by going over image subsetting. There were two ways in which we did this. the first was with an inquire box. The second way was with an area of interest shape file. The following images are the result of part 1 of our lab.
In part two of lab four, we learned image fusion. We started with a coarse resolution image and changed the spatial resolution to make our final image more appealing. We utilized pansharpening to make the image easier to view at a large scale. We did this by combining our image with its panchromatic counterpart which has a higher resolution. The end result was a more clear version of our original image.
Part three of the lab was where we learned simple radiometric enhancement technique. This is a useful technique to reduce haze that will sometimes appear in satellite images. It allows us to view the image clearly and completely.
In our fourth part of lab four, we learned how to link an image viewer to Google Earth. This was very interesting to work with. It made it easier to compare the land and the image. We also learned in this section that Google Earth can be considered a selective key.
In part five of lab four, we worked with resampling. We resampled an image using both the nearest neighbor method as well as the bilinear interpolation method. When we compared the nearest neighbor resampled image to the original image, there wasn't much of a difference. However, when we observed the differences in he bilinear resampled image, we could see a definite improvement in visual appeal.
In part six of lab four, we worked with image mosaicking. We needed to combine two images that stacked and connected to each other. the first time we used Mosaic Express. This produced an image that could clearly tell where the two images were combined. The second method we used was MosaicPro. This method allowed for much more customization. It took longer to produce, but the end result was a much higher quality mosaic image. The following images are the results from part six.
Part seven of lab four taught us binary change detection. This is otherwise known as image differencing. In section one we created a difference image. For this we used the "Two Input Operators" interface. When viewing the histogram, we could figure out where the points that differed from the original image were. In the second part, we used a formula to map out the pixels that changed instead of creating an image to show it. The following images are the results of part one and two of part seven.
