Lab 8: Photogrammetry
Introduction & Goal
In this lab, the objective is to develop skills in performing photogrammetric tasks on vertical aerial photographs and satellite images. Overall, this lab will help students develop skills and understanding the mathematics behind the calculation of photographic scales, measurement of areas and perimeters of features, calculating relief displacement, and performing orthorectification on a blog of vertical aerial photographs.
Methods
For the first part of this lab, JPEG images were used to calculate photographic scales when measuring real world areas, topographic relief displacement, and perimeters. Firstly, the image Eau Claire_West-se.img (an aerial photograph) was used to determine the distance between points A and B on the JPEG photograph to overall determine the scale. Because the given distance is 8822.47 ft, you must convert this to inches, and measure the distance from A-B with a ruler (inches). Afterwards, utilize the scale equation: Scale (S) = Photo distance (P) - Ground Distance (G) to calculate the scale of the image. Then, we implemented this method to the next image, ec_east-sw.img. I utilized this equation: Scale (S) = Focal Length of camera (F) / [Flying Height above sea level(H) - City elevation above sea level (h)] with h being the elevation of Eau Claire (S=F/(H-h)) to calculate the scale of this photo.
Next, we focused on measurement of areas and features on an aerial photograph. Utilizing the polygon tool on the aerial image ec_east-se.img (NAIP 2005 image) in the viewer, we utilized the Polygon tool to digitize the outline of a lagoon. To obtain the perimeter and area of the measurements, read directly from the View Measurements table at the bottom of the image. You can change the units as well to see your data in hectares, miles, acres, meters, etc. Also, we identified and calculated relief displacement of tall objects in the images with the equation: D = (h x r)/H. I chose to look at both the water tower on the south side of Eau Claire and the smoke stack near Sacred Heart Hospital to identify which image had distinct relief displacement. When approximating, the relief displacement of the smoke stack is +0.26".
For the second part of this lab, we focused on Orthorectification where we orthorectified three raw vertical aerial photographs collected from southeast Spain to remove horizontal (x,y) and vertical (z) errors. These images have their spatial resolution of 0.4 meters and the sensor that was used is a Kodak DCS 420 digital camera. In this section of the lab, we created a photogrammetric project, added imagery to the block file, defined the camera model, performed automatic tie point collection, performed triangulation, orthorectified the images, and viewed said orthorectified images.
Before creating a photogrammetric project, we displayed the raw photographs collected on this airborne flight mission. We then proceeded to fix the images associated with the photos by orthorectification. We created a New Block File window and found under Camera (Geometric Model Category), we selected Digital Camera. We then set the projection of the image under the Block Property Setup Dialog box for the Horizontal Reference Coordinate System. Using WGS84 --> WGS84 --> UTM Zone 30 --> North --> E,N --> and EPSG Code: 32630, the reference coordinate system reflects the projection selected. We left the Vertical Reference Coordinate System at WGS 84 default. In the Set-Frame-Specific information window, the Average Flying Height (meters) were taken at 1248.168 meters. The average flying height is an average of the principal point elevation for each of the photographs taken.
When clicking on the import exterior orientation parameters button, this opens The Import ASCII Name dialog box. Here, the Import Parameters dialog opens, and you can then verify if the map's information matches the specifications you previously entered in the Block file. After clicking okay, click on Row Terminator Character dropdown list and select Return Newline (DOS) from the dropdown menu. Preview your input, and click OK. I noticed that the image names with their x, y, z, Omega, Phi, and Kappa values are noted as well and imported. Now, the Photogrammetry module is open!
In order to proceed, the images from Spain need to be online which includes attaching and linking the image name to the corresponding image. When clicking on a cell, the Digital Camera Frame Editor dialog opens where you can click the attach button and navigate to and select the correct TIFF files. Returning to Photogrammetry, the cells in the dialog box are green and ready to go. We then computed pyramid layers for the images in the block file, which are essential for optimizing image display and tie-point generation. You can do this by clicking on the first cell below Pyr column to calculate the first image's pyramid layers. Do the same for image two and when the dialog box shows up, confirm that all images without pyramids is checked.
Next we proceeded to perform interior orientation by clicking on the first cell in the Int column to open the Digital Camera Frame Editor dialog box. By clicking on the New Camera button, we entered the Camera Information used to obtain the images in Spain. After confirming the Principal Point xo and Principal Point yo values and clicked save, the camera parameter file name dialog opens to navigate to the lab folder and from there can be saved. After clicking next, confirming that the Sensor Name dropdown list shows Kodak DCS 420 Digital Camera, return to digcam1 to enter Pixel Size information and apply all to active frames. Click the next button twice to ensure that the same interior orientation information is transferred over to digcam2 and 3.
We then performed exterior orientation where we clicked the Exterior Orientation tab in the Digital Camera Frame Editor dialog box. By clicking the next button twice the information values for the third and second image will show up for digcam2 and digcam 3. After hitting Ok, we ended up back in the Photogrammetry suite.
Figure 1) Exterior Orientation finished and the Photogrammetry suite is visible.
To determine the appropriate parameters for exterior orientation, we needed to perform automatic tie points generation. By clicking the Point Measurement tool, it opens up to display three views, a Main View and two Cell Arrays. Click the Automatic Tie Properties icon to open up the generation properties in the general tab. In the distribution tab, type 50 in the intended number of points per image filed. Click Run to display the tie point information for each image in the viewer and review the Auto Tie Summary dialog before saving in the Point Measurement tool.
Figure 2) Point Measurement Tool activated before displaying tie-point information.
Since we have now generated the tie points, we can now proceed with triangulation. Click the first cell under EXT to open the Aerial Triangulation properties dialog to compute the accuracy of the adjusted exterior orientation parameters and the x, y, and z coordinates. By altering the coordinate units and weight value for x, y, z and confirming that omega, pi, and kappa fields are set to 0.1, we can remove poor tie points from the solution. When the triangulation summary dialog opens, make sure all parameters are correct and then when the triangulation report opens, this report can be saved as a text file for future reference. We reviewed the triangulation report to find and remove the points with the most error before Accepting the dialog.
Lastly, we created orthorectified images by executing Orthoresampling. The latter produces orthorectified images that are planimetrically sound (the removal of horizontal and vertical errors). Create a new Photogrammetric project by creating a new block file. After entering the parameters described in the lab in the ortho resampling dialog box, we chose the resampling method is Bilinear Interpolation. By clicking the add multiple button, we can ortho sample multiple images at the same time and automatically adds the ortho output file. When observing the cells under Ortho, they should all be green except for DTM, which signifies that the Orthoresampling process is complete and can be brought into a new viewer to display the orthorectified images.
Results
Figure 4) Screen Capture of the Orthorectified Images.
Data Provided by:
- UWEC Geography Department
- National Agriculture Imagery Program (NAIP) images are from United States Department of Agriculture, 2005.
- Vertical aerial photographs are from Erdas Imagine, 2016.
- GPS data is from Erdas Imagine, 2016.
Comments
Post a Comment