Lab 7: Stereoscopy and Orthorectification

Goals and Background:

The purpose of this lab was to gain skills in photogrammetry using aerial and satellite images. The first part of the lab developed an understanding of how to calculate scales, measure area and perimeters of features, and calculate relief displacement. The rest of the lab developed skills in stereoscopy and orthorectification of satellite images. The lab helped me develop photogrammetry skills, which will be useful in a remote sensing career.

Methods:

Part 1: Calculating scale of nearly vertical aerial photographs

Section 1:

The scale of an aerial photograph (Eau Claire_West-se.jpg) of Eau Claire (Erdas Imagine, 2009). was calculated using the equation below:

S = pd/gd  (1)

Where:
S = scale
pd = distance on the map
gd = distance in the real world

The distance on the map was measured using a ruler and the ground distance was provided (8822.47 ft).

The scale of a second aerial photograph, collected by a high altitude reconnaissance aircraft (United States Department of Agriculture, 2005), was calculated using the equation below:

S = f/(H-h)  (2)

Where:
S = scale
f = focal lens length
H = altitude above sea level
h = elevation of terrain

The focal lens length (152 mm), altitude above sea level (20,000 ft), and elevation of terrain (796 ft for Eau Claire) was provided.

Section 2: 

The area and perimeter of a lagoon in an aerial photograph was measured using the "Measure Perimeters and Area" tool in Erdas Imagine 2013. Area was reported both in ha and acres, whiles perimeter was recorded both in meters and miles.

Section 3:

The relief displacement of a smoke stack in an aerial photograph of Eau Claire was calculated using the below equation:

d = (h*r)/H  (3)

Where;
d = relief displacement
h = height of object (real world)
r = radial distance of top of displaced object from principal point (photo)
H = height of camera above local datum

The height of the camera above the local datum was 3980 ft,  The real world height of the smoke stack and the radial distance were calculated using a ruler and the scale of the aerial photograph (1: 3209).


Part 2: Stereoscopy

Ground control points (GCPs) were used to create a 3D display of the City of Eau Claire (Figure 1). A DEM (ec_dem2.img) was used to create the stereoscopic view of Eau Claire (ec_city.img) using the Anaglyph tool.

Part 3: Orthorectification

The Erdas Imagine Lecia Photogrammetric Suite (LPS) was used to orthorectify and planimetrically correct a satellite image of Palm Springs, CA (Figure 2).

Section 1:

A new block file was created in LPS Project Manager using SPOT satellite images of Palm Springs, CA (Erdas Imagine, 2009). The geometric model interface was set to "polynomial-based pushbroom" and the geogmetric model category was set to "SPOT pushbroom". In the "Projection Chooser" dialog (in Horizontal Reference Coordinate System section of the block Property Setup dialog) the following parameters were set:

Projection Type: UTM
Spheroid Name: Clarke 1866
Datum name; NAD27(CONUS)
UTM Zone: 11
North or South field: North
Horizontal Units: Meters

No changes were made to the Vertical Section of the reference Coordinate System.

Section 2: 

Satellite imagery of Palm Spring, CA (Spot_pan.img) was added to the block and the parameters of the SPOT pushbroom sensor were verified during this section of the lab. No changes needed to be made to the image, but the verification process was carried out so that the sensor would be specified. The verification processed turned the Int. column in LPS Project Manager to green, which would later allow me to complete the orthorectification process.

Section 3:

GCPs were collected for the image spot_pan.img using the Classic Point Measurement Tool. An orthorectified image (sx_ortho.img) was used for collecting GCPs. The GCPs were collected by first selecting points on the reference image (sx_ortho.img). Reference control points had to match provided control point coordinates within 10 meters (Table 1). Corresponding reference coordinates were collected on the spot_pan.img using the Point Measurement Tool. Coordinates had to match provided control point coordinates within 2 pixels (Table 2). 

Point ID	X Reference	Y Reference
1	566189.190	3773586.979
2	555690.659	3728387.770
3	501918.953	3732595.411
4	515114.084	3759740.576
5	543537.306	3779981.255`
6	558640.300	3751516.718
7	532062.982	3724946.633
8	539381.670	3768419.388
9	526013.661	3753709.856
11	545372.750	3741643.250
12	540901.659	3746876.633

Table 1

Point ID	Image Name	X File	Y File
1	Spot_pan	5239.468	337.384
2	Spot_pan	5191.590	1969.546
3	Spot_pan	230.925	5378.823
4	Spot_pan	869.542	2487.996
5	Spot_pan	3027.570	51.432
6	Spot_pan	4999.412	2636.848
7	Spot_pan	3064.254	5673.794
8	Spot_pan	2890.880	1258.852
9	Spot_pan	1978.138	2919.004
11	Spot_pan	3982.969	3817.813
12	Spot_pan	3469.092	3367.939

Table 2

After the first two GCPs were collected, the Automatic (x.y) drive function was used to help collect GCPs faster. Using the Automatic (x,y) drive, coordinate values were input into the Measurement Point Tool, which placed the coordinate on the reference image and positioned the regular image in the approximate location of the coordinate.

The last two control points (Point ID 11 and Point ID 12), were collected from NAPP_2m-ortho.img, a different horizontal reference source. The 10th control point was skipped to make the difference between the two different horizontal reference sources clear.

Elevation information from the DEM file titled palm_springs_dem was collected for the horizontal reference GCPs obtained from the xs_ortho and NAPP_2m-ortho images. This task was accomplished using the Reset Vertical Reference Source icon in the Point Measurement tool palette. The Update Z Values on Selected Points icon was used to update z values of all reference points in the cell array based on the values in the palm_springs_dem.

Section 4:

The Type and Usage values for the control points were set using the Formula dialog in Point Measurement Tool. All type values were set to Full, while all Usage values were set to Control.

A second image (spot_panb.img) was added to the block because the collection of reference points on the first image was complete. Spot Pushbroom Frame Editor parameters were accepted, which tuned the Int column in the LPS Project Manager to green.

GCPs were collected on spot_panb.img based on the points already collected in spot_pan. Collected GCPs for the spot_panb.img had to match within two pixels of the values in Table 3. Point IDs 3, 4, and 7 were not collected for spot_panb.img becase they were not on the image.

Point ID	Image Name	X File	Y File
1	Spot_panb	2857.270	753.852
2	Spot_panb	3003.782	5387.892
5	Spot_panb	1022.701	644.456
6	Spot_panb	2736.125	3070.227
8	Spot_panb	937.482	1862.696
9	Spot_panb	221.445	3594.113
12	Spot_panb	1499.230	3923.753

Table 3

Section 5:

Tie points were collected for spot_pan.img and spot_panb.img using the Automatic Tie Point Generations Properties icon in the Point Measurement tool palette. The following parameters were used:

Image used: All Available
Initial Type: Exterior/header/GCP
Image Layer Used for Computation: 1
Intended Number of Points/Image: 40
Keep All Points: check box is unchecked

The accuracy of a few tie points were confirmed to ensure proper orthorectification.

Triangulation was performed using the LPS Project Manager. The following parameters were used:

Iterations With Relaxation value: 3
Image Coordinate Units for Report: Pixels
Ground Point Type: Same Weighted Values
Ground Point Standard Deviations: 15 (for X, Y, and Z fields)
Simple Gross Error Check Using: check box is checked
Times of Unit Weight: 3.0


The triangulation report was saved as a ASCII Text file. It should be noted the Ext. columns in the cell array were green after completing the triangulation process.

The final orthorectified image was created using Start Ortho Resampling Process icon in LPS Project Manager. The following parameters were used:

DTM Source: palm_springs_dem.img
Output Cell Sizes: 10.0 (for both X and Y)
Resampling method: Bilinear Interpolation
Input File Name: spot_panb
Use Current Cell Sizes: check box is checked

Section 6:

The orthorectified images were viewed using LPS Project Graphic Status and an image viewer (Figure 2). The swipe function was used to evaluate the spatial accuracy of the area of overlap in the images. The block file was saved before closing the program.

Results:

Figure 1: Anaglyph image of Eau Claire shows changes in elevation with 3D.

The elevation changes are most pronounced around Putnam Park at the center of the image,

Figure 2: The orthorectified images of Palms Springs, CA, is relatively seamless.

Sources:

Hexagon Geospatial. (2009). Erdas Imagine [computer software]. Georgia: Norcross. 

United States Department of Agriculture. (2005). [Satellite images in img. format]. National Agriculutre Imagery Program. Retrieved from: https://gdg.sc.egov.usda.gov/.

United States Department of Agriculture Natural Resources Conservation Service. (2010). [Digital elevation model for Eau Claire in dbf. format]. Retrieved from: http://www.nrcs.usda.gov/wps/portal/nrcs/site/national/home/.

Lab 6: Geometric Correction

Goals and Background:

The goal of the lab is to develop knowledge and first-hand skills in geometric correction. Geometric correction is an image preprocessing exercise that is used on satellite images prior to any data extraction from the satellite images. Skills learned in the lab will be useful to future work with remotely sensed images.

Methods:

Part 1:

Geometric correction of the Chicago_2000.img Landsat TM image (United States Geological Survey, 2015) was performed using a United States Geological Survey 7.5 minute digital raster graphic of the same area (Illinois State Geological Survey ,2015). This was accomplished by adding ground control points (GCP's) on both the input and reference images using the Multipoint Geometric Correction dialog box in ERDAS IMAGINE 2013. A first order polynomial equation was used to rectify the image. Four GCP's were added to each image and a total root mean square (RMS) error of 2 or lower was achieved by replacing the GCP's in more accurate locations. The Multipoint Geometric Correction dialog box with an appropriate RMS error can be seen in Figure 1. The geometric correction was performed after the RMS error was lowered to 2 or less. The Display Resample Image Dialog button on the Multipoint Geometric Correction tool bar was used to complete the geometric correction process. The Chicago image was geometrically corrected using nearest neighbor interpolation and accepting the other default values. The results of the geometric correction for the Chicago image can be seen in Figure 2.

Part 2:

Geometric correction of the geometrically distorted Sierra_Leone_east1991.img   was performed using a corrected Landsat TM reference image of the same area called Sierra_Leone_east1991grf.img (United States Geological Survey, 2015). The correction process was identical to the process carried out in Part 1 except for a couple exceptions. First, a third polynomial equation was used to correct the image.  Second, twelve GCP's were used. Third, an RMS error of 1 or lower was achieved to correct the image (Figure 3). Finally, bilinear interpolation was used when completing the geometric correction process. The results of the geometric correction for the Sierra Leone image can be seen in Figure 4.

Results:

Figure 1: Total RMS error for Chicago_2000.img is less than 2.0.

Figure 2: Geometrically corrected Chicago image using nearest neighbor interpolation.

Figure 3: Total RMS error for Sierra_Leone_east1991.img is less than 1.0.

Figure 4: The geometrically corrected image for Sierra Leone, which used bilinear interpolation, is relatively spatially accurate.

References:

Illinois State Geological Survey. (2015). [Digital raster graphic is in img. format]. Illinois Geospatial Data Clearing House. Retrieved from http://www.isgs.uiuc.edu/nsdihome/

United States Geological Survey. (2015). [Satellite images are in img. format]. Earth Resources Observation and Science Center. Retrieved from http://eros.usgs.gov/ 

Lab 5: Creating a Digital Surface Model and Digital Terrain Model from Lidar Data

Goals and Background:

The purpose of the lab was to gain experience in using lidar data. An LAS dataset was explored with ArcMap 10.2.2, and lidar derivative products, such as DSM and DTM, were created. Lidar is a growth industry that has boomed recently due to its useful applications. Skills learned in the lab will serve students well in the workforce.

Methods:

Part 2:

An LAS dataset was created using ArcMap 10.2.2 and files were provided by Eau Claire County (2013). Features of the LAS dataset were explored by viewing the statistics of the data. Statistics were used for a quality check of the data to determine if elevations in the data were accurate to real world data.  Additionally, a shapefile of Eau Claire County (Price, 2014) was added to the LAS data set to determine if the data was spatially correct. The horizontal coordinate system was assigned to NAD 1983 HARN Wisconsin CRS Eau Claire (US Feet), and the vertical coordinate system was assigned NAVD 1988 US feet.

Point clouds were visualized in ArcMap 10.2.2 using the LAS Dataset Profile View tool for a 2D representation of any location in the dataset. The LAS Dataset 3D View tool was used to see the same location in 3D.

Part 3:

Section 1:

A digital surface model (DSM) and digital terrain model (DTM) were created using the LAS Dataset to Raster Tool in ArcMap 10.2.2. The results for the DSM and DTM can be seen in Figures 1 and 3, respectively.

The parameters for the DSM model were:

• Interpolation: Binning
• Cell Assignment Type: Maximum
• Void Filling: Natural Neighbor
• Sampling Type: Cellsize
• Sampling Value Field: 6.56168ft (2m)

The parameters for the DTM model were:

• Interpolation: Binning
• Cell Assignment Type: Minimum
• Void Filling: Natural Neighbor
• Sampling Type: Cellsize
• Sampling Value Field: 6.56168ft (2m)

Hillshades were created for the DSM and DTM images using the Hillshade tool under 3D Analyst Tools. The hillshades for the DSM and DTM can be seen in Figures 2 and 4, respectively.

Section 2:

A lidar intensity image of the Eau Claire area was created using the LAS Dataset to Raster tool. Before the tool was opened, the LAS Dataset was set to Points and the Filter was set to First Return. Once in the LAS Dataset to Raster tool, the following parameters were used:

• Interpolation: Binning
• Value Field: Intensity
• Cell Assignment Type: Average
• Void Filling: Natural Neighbor
• Cell Size: 6.56168ft (2m)

After the image was created in ArcMap 10.2.2, it was loaded into ERDAS IMAGINE 2013 to enhance the display of the image. The intensity image can be seen in Figure 5.

Results:




Figure 1: DSM of Eau Claire

Areas of lower elevation are darker and areas of higher elevation are lighter in the DSM.



Figure 2: Hillshade of Eau Claire DSM

The hillshade of the DSM shows objects on the ground clearly. 

Figure 3: DTM of Eau Claire

The DTM is a bare-earth representation of the Eau Claire area. 

Figure 4: Hillshade of Eau Claire DTM

The hillshade of the DTM does not have any ground objects and is useful for interpreting the bare landscape.

Figure 5: Intensity image of Eau Claire

The intensity image records the strength of the return, and is useful for interpreting lidar data.

References:

Eau Claire County. (2013).

Price, M. (2014). Mastering ArcGIS 6th Edition. Mastering ArcGIS 6th Edition Dataset [shapefile]. New York: McGraw Hill.