The orthoimagery in this image service was taken on 7/21/1951 over Harrisonburg, VA. Data was supplied from James Madison University. Dr. Glen Gustafson's ISAT Applied GIS Lab created this mosaic for the City of Harrisonburg from scanned prints. The dataset's coordinate system is WGS 1984 Web Mercator (Auxiliary Sphere). A cache was built using the the ArcGIS Online/Bing Maps/Google Maps tiling scheme containing 11 levels ranging from 1:288,895 to 1:282.

Metadata from JMU student

Matthew Chenault - May 6, 2004

Harrisonburg, VA Orthophoto Production and Mosaic - Geog 490, Independent Study

The goal of this assignment was to orthocorrect and mosaic four panchromatic air photos taken on July 21, 1951 of the City of Harrisonburg and areas of surrounding Rockingham County. The exposures used in this exercise are identified as DJN-12G-102, DJN-12G-103, DJN-12G-104 and DJN-12G-105.

The first step of the project was to scan each raw airphoto at 600dpi and import it into ERDAS Imagine. With this newly created .img file I followed the steps of defining the DEM to be used for elevation, entered the camera focal length of 208.44mm and digitized the fiducial marks on the airphoto. I then entered the corresponding map coordinates with these digitized fiducials and set the projection of the soon to be created orthophoto to match the provided reference map. I orthocorrected each imported air photo by georeferencing it with the mosaicked 1987 Harrisonburg/Bridgewater USGS 2.43m DRG. I specified the mosaicked Harrisonburg/Bridgewater 30m USGS DEM to correct for elevation and resampled each photo with an output cell size of 0.75 meters. I initially used approximately 40 GCP points on each photo and then deleted several of them to reduce the RMSE so that each photo had an RMSE of less than 6 pixels. After creating each orthophoto I ensured that it matched accurately with the reference map projection and that it had the appropriate cell size. After performing the initial quality control on each orthophoto, I began the mosaic process using the Data Prep module in ERDAS. After adding each image to the viewer, I selected “Compute Active Area” and chose to crop 30% off of each orthophoto during the mosaic process. The 30% was the minimum necessary percentage to crop the photo to fully remove any remnants of the photos’ date and exposure number. I selected “Bilinear Interpolation” as the resample method and “No Matching” under the matching options menu. I then selected “Feather” under the Select Overlap Function dialogue and proceeded to the Output Image Options menu. There I selected “Union of All Inputs” and specified an output cell size of 0.75 meters and an output data type of Unsigned 8-bit. After specifying the output file name I selected “Ignore Zeros” in the mosaic dialogue and then ran the mosaic process. I chose to Subset the initial mosaicked image using the Data Prep module by selecting an AOI of the mosaic that created a rectangle. I specified the new output name, the matching data type and cell size to produce the subset.

The visual quality of the output images is very comparable to the input images. There is no noticeable difference in the brightness of the areas covered by the photos and there do not appear to be any major blunders or errors in the data. Most importantly the line between the photos is undetectable. The image information window shows that this data is indeed in the correct projection and contains the correct pixel size as specified in the mosaic process. I measured the horizontal accuracy of each orthophoto and the mosaic by swiping with the reference map and measuring prominent intersections. It is important to note here that many of the roads in the City may have shifted or expanded since 1951 and there naturally may be some discrepancy with the DRG. Therefore I only used prominent intersections that appeared to exist during both time periods. I observed that these intersections varied from the DRG by 0 to 6 meters. I also obtained the City Street Centerline shapefile from Sam Hottinger to measure the accuracy of the photos with some sample data from the City. The prominent intersections and streets in the orthophotos varied from the Centerline file by 0 meters to 6 meters, generally in the southwest direction. This was a very interesting assignment because it compiled many of the techniques learned in remote sensing into one comprehensive project. It is very rewarding to serve the needs of the City and hopefully they will be able to put our work to good use.

The orthoimagery in this image service was taken on 7/21/1951 over Harrisonburg, VA. Data was supplied from James Madison University. Dr. Glen Gustafson's ISAT Applied GIS Lab created this mosaic for the City of Harrisonburg from scanned prints. The dataset's coordinate system is WGS 1984 Web Mercator (Auxiliary Sphere). A cache was built using the the ArcGIS Online/Bing Maps/Google Maps tiling scheme containing 11 levels ranging from 1:288,895 to 1:282.

Metadata from JMU student

Matthew Chenault - May 6, 2004

Harrisonburg, VA Orthophoto Production and Mosaic - Geog 490, Independent Study

The goal of this assignment was to orthocorrect and mosaic four panchromatic air photos taken on July 21, 1951 of the City of Harrisonburg and areas of surrounding Rockingham County. The exposures used in this exercise are identified as DJN-12G-102, DJN-12G-103, DJN-12G-104 and DJN-12G-105.

The first step of the project was to scan each raw airphoto at 600dpi and import it into ERDAS Imagine. With this newly created .img file I followed the steps of defining the DEM to be used for elevation, entered the camera focal length of 208.44mm and digitized the fiducial marks on the airphoto. I then entered the corresponding map coordinates with these digitized fiducials and set the projection of the soon to be created orthophoto to match the provided reference map. I orthocorrected each imported air photo by georeferencing it with the mosaicked 1987 Harrisonburg/Bridgewater USGS 2.43m DRG. I specified the mosaicked Harrisonburg/Bridgewater 30m USGS DEM to correct for elevation and resampled each photo with an output cell size of 0.75 meters. I initially used approximately 40 GCP points on each photo and then deleted several of them to reduce the RMSE so that each photo had an RMSE of less than 6 pixels. After creating each orthophoto I ensured that it matched accurately with the reference map projection and that it had the appropriate cell size. After performing the initial quality control on each orthophoto, I began the mosaic process using the Data Prep module in ERDAS. After adding each image to the viewer, I selected “Compute Active Area” and chose to crop 30% off of each orthophoto during the mosaic process. The 30% was the minimum necessary percentage to crop the photo to fully remove any remnants of the photos’ date and exposure number. I selected “Bilinear Interpolation” as the resample method and “No Matching” under the matching options menu. I then selected “Feather” under the Select Overlap Function dialogue and proceeded to the Output Image Options menu. There I selected “Union of All Inputs” and specified an output cell size of 0.75 meters and an output data type of Unsigned 8-bit. After specifying the output file name I selected “Ignore Zeros” in the mosaic dialogue and then ran the mosaic process. I chose to Subset the initial mosaicked image using the Data Prep module by selecting an AOI of the mosaic that created a rectangle. I specified the new output name, the matching data type and cell size to produce the subset.

The visual quality of the output images is very comparable to the input images. There is no noticeable difference in the brightness of the areas covered by the photos and there do not appear to be any major blunders or errors in the data. Most importantly the line between the photos is undetectable. The image information window shows that this data is indeed in the correct projection and contains the correct pixel size as specified in the mosaic process. I measured the horizontal accuracy of each orthophoto and the mosaic by swiping with the reference map and measuring prominent intersections. It is important to note here that many of the roads in the City may have shifted or expanded since 1951 and there naturally may be some discrepancy with the DRG. Therefore I only used prominent intersections that appeared to exist during both time periods. I observed that these intersections varied from the DRG by 0 to 6 meters. I also obtained the City Street Centerline shapefile from Sam Hottinger to measure the accuracy of the photos with some sample data from the City. The prominent intersections and streets in the orthophotos varied from the Centerline file by 0 meters to 6 meters, generally in the southwest direction. This was a very interesting assignment because it compiled many of the techniques learned in remote sensing into one comprehensive project. It is very rewarding to serve the needs of the City and hopefully they will be able to put our work to good use.