There are several types of raster data that may be used as a source for feature extraction. Some may originate from satellite imagery, flown imagery, aerial photography, or a scanned image of an existing map or chart. Regardless of origin, the raster must be accurately georeferenced to be useful.
All raster datasets are in some coordinate space. This coordinate space may be the real-world coordinate system or image space. Since almost all raster datasets represent some real-world location, it is best to have that dataset in a real-world coordinate system that best represents it. Converting a raster dataset from a non-real-world coordinate system (image space) to a real-world coordinate system is called georeferencing.
Typically, in DNC, a chart has been scanned and needs to have real-world coordinates applied to it for extraction or maintenance purposes. To attach the real-world coordinates, you first add control points to the scanned chart, then rectify the image. Before you begin, investigate either the hard- or soft-copy chart and identify locations where the meridians and parallels intersect; these intersections are what you use to perform the referencing. You may want to copy the coordinates and convert them to decimal degrees before you begin georeferencing as it saves you some time to have all possible coordinates in a table you can reference easily.
Georeferencing images with existing spatial information
Occasionally, the National Geospatial-Intelligence Agency (NGA) provides images that have already been georeferenced. In this case, use the project on-the-fly capabilities of ArcMap.
- Load the DNC geodatabase layers you are going to be working on
- Load the image
By loading the geodatabase layers, the coordinate system for the data frame is set to geographic coordinates in the WGS 1984 datum because that is the information attached to the database. When you choose to load an image with a different projection or datum, ArcMap recognizes it and projects the image to the coordinate system of the data frame. You can check the data frame coordinate system by opening the Data Frame Properties dialog box.
If you were to open the image first and then bring in the geodatabase layers, you can easily change the coordinate system to GCS WGS 1984. It is always important to do new collection, corrections, maintenance activities, and other tasks in the native units of the geodatabase you're working with. There are many times throughout these activities that you may want to switch projections (QC, Notice to Mariners), but always be sure to return the data frame coordinate system to WGS 1984 once that task is complete.
Applying a shift
You can use one of the following methods to apply a shift to a chart:
- Apply the shift to the coordinates of the tick marks when georeferencing the image. With each coordinate pair entered, take into account the shifts necessary to bring the data to the WGS 1984 datum.
- Do not apply the shifts when georeferencing the data. Enter the coordinates exactly as shown on the scanned chart.
With the second method, when defining the spatial reference of the image, you need to choose the geographic projection, but then choose the appropriate datum for the coordinates entered. This way, ArcMap recognizes the image is in a different datum and projects it for use during production.