Goal and Background
The main goal of this laboratory exercise was to develop my skills in performing key photogrammetric tasks on aerial photographs and satellite images. Specifically the lab was able to help me understanding the mathematics behind the calculation of photographic scales, measurement of areas and perimeters of features, and calculating relief displacement. Moreover this lab was indented to introduce me to stereoscopy and performing orthorectification on satellite images.
Methods
First, I performed basic scale, measurement and relief displacement calculations using various photogrammetric equations. In the first exercise I calculated the scale of a photograph using a real world distance and a on-photo measured distance. Next I calculated the scale of the photograph with another equation that uses altitude, focal length, and elevation. I continued by using Erdas Imagine's 'Measure Perimeters and Areas' digitizing tool to calculate the area and perimeter of a body of water in the Eau Claire area. Next, I calculated relief displacement using object height, principal point, and elevation.
Next, I used Erdas Imagine's 'Anaglyph Generation' tool to create an anaglyph photo from an image of Eau Claire and a digital elevation model (DEM) of the city of Eau Claire at 10 meters spatial resolution. Noting that I set the vertical exaggeration to 2 (rather than 1), I analyzed the anaglyph image's features in relation to real life features.
Finally, I performed orthorectification using Erdas Imagine Lecia Photogrammetric Suite (LPS). First, I created a new project in LPS project manager, and created a new block file. I set the projection system according to my images proceeded to add a frame which subsequently contained an image. I verified that the sensor properties matched my image's appropriate sensor settings (SPOT). Next, I activated the point measurement tool to collect ground control points (GCPs). I used the horizontal reference tool to activate my reference image that facilitates the collection of GCPs. I used an already orthorectified image of Palm Springs, CA as a reference image, and began collecting GCPs. I did this by placing a point on the reference image, and one in the same geographic location on the un-rectified image. The (x,y) coordinates were provided to facilitate GCP collection. I collected 9 GCPs using this method. I then reset the horizontal reference source and collected two more ground control points using a different ortho-image. I then assigned z values to all of the GCPs using a DEM image of the Palm Springs area. Next I changed the GCP's type to 'Full' and their usage to 'Control' to specify their uses and properties. I then returned to the LPS project manager to add another frame and another image. Again, I verified the sensor properties and reopened the point measurement tool. In the new image, I began collecting GCPs that corresponded to those in the first image. The points that were not present on the second image were skipped because they did not overlap. Next, I edited the automatic tie point generation properties to fit my purposes, and created tie points based on the already present ground control points. Some 25 were automatically generated. I then edited the Triangulation properties in LPS project manager to establish the mathematical relationship between the images that make the block file, the sensor model, and the ground. I ran the model, viewed the report, and accepted the triangulation results. I finally selected the 'Start Ortho Resampling Process' icon in LPS project manager and verified my DEM settings, output file name, resampling method and made sure both project images were included. I viewed the results of the orthorectified images, noting that the spatial match between them was very accurate.
Results
 |
| Ortho-rectified overlapping images of Palm Springs, CA area |
Sources
None
