Remote Sensing Lab 4

Goal and Background - This lab was focusing on image subset for finding an area of interest (AOI), haze reduction on satellite images, Google Earth, and resampling of images to improve resolution. All of this within the program Erdas Imagine 2013.

Methods and Results - For the first part of the lab we learned how to create a subset image using the inquire box and the subset and chip tool. I first selected the AOI using the inquire box tool; then, using the subset and chip tool I created a subset image from the inquire box (see fig. 1).

fig. 1

The next section was the same as the first but instead using a shapefile as a AOI. I first took a shape file of a specific AOI and put it on top of a satellite image. I then used the paste from selected object tool to select the area of the satellite image that lies in the AOI shapefile. Next, I used the subset and chip tool from the previous section to create a subset image of the shapefile AOI (see fig. 2).

fig. 2

Part two of the lab was about image fusion. I took a reflective image with a spatial resolution of 30 meters and using the resolution merge tool I merged the original image with a pansharpened image with a 15 meter spatial resolution. This created a new pan-sharpened image with higher spatial resolution.

Part three was focused on radiometric enhancement techniques. I took a image and using the haze reduction tool I reduced the haze making the image clearer and more colorful.

Part four was about linking the image viewer in Erdas to Google Earth. I opened Google Earth in Erdas by clicking connect to Google Earth under the Good Earth tab. I then opened an image in the image viewer and clicked match GE to view and then sync GE to view. This can be used to interpret images since the Google Earth image has such high spatial resolution and 3D capabilities.

The final part was on resampling. I took a image with 30 meter spatial resolution and used two different resampling techniques to enhance the resolution and lower the pixel size. The first technique was nearest neighbor with when used it creates a pixel pattern (see fig. 3). The second technique was bilinear intepolation which resampled the pixels into smaller 20 meter resolution uniform pixels giving us a more detailed image.

fig 3

Sources - All images provided by Professor Wilson.