Geo referencing in GIS
What is Georeferencing? 🗺️
Georeferencing is the process of aligning a dataset, such as a scanned map or an aerial photograph, to a known geographic coordinate system. It gives a raster image spatial context by assigning real-world coordinates to every pixel. This allows data without spatial information to be used and analyzed alongside other georeferenced data within a GIS. Without this process, a scanned map is just a picture; with it, the map becomes a usable layer for analysis.
Georeferencing is the process of aligning a dataset, such as a scanned map or an aerial photograph, to a known geographic coordinate system. It gives a raster image spatial context by assigning real-world coordinates to every pixel. This allows data without spatial information to be used and analyzed alongside other georeferenced data within a GIS. Without this process, a scanned map is just a picture; with it, the map becomes a usable layer for analysis.
Georeferencing vs. Geocoding
Georeferencing gives a raster image spatial context by aligning it to a coordinate system.
Geocoding converts text-based addresses (e.g., "123 Main Street") into geographic coordinates.
Georeferencing gives a raster image spatial context by aligning it to a coordinate system.
Geocoding converts text-based addresses (e.g., "123 Main Street") into geographic coordinates.
Foundational Concepts and Key Components
Key Concepts
Spatial Data: Information about the location and shape of geographic features, such as points, lines, or polygons.
Raster Data: A type of spatial data represented as a grid of pixels, often used for images like aerial photos or satellite imagery.
Vector Data: Represents features using points, lines, and polygons, such as roads, rivers, and boundaries. Vector data is already spatially defined and does not require georeferencing.
Coordinate System: A system that uses numbers to define locations on the Earth's surface. Common types include geographic coordinates (latitude and longitude) and projected coordinates (like UTM).
Reference Data: An accurate map or spatial dataset with known coordinates used as a base for alignment.
Spatial Data: Information about the location and shape of geographic features, such as points, lines, or polygons.
Raster Data: A type of spatial data represented as a grid of pixels, often used for images like aerial photos or satellite imagery.
Vector Data: Represents features using points, lines, and polygons, such as roads, rivers, and boundaries. Vector data is already spatially defined and does not require georeferencing.
Coordinate System: A system that uses numbers to define locations on the Earth's surface. Common types include geographic coordinates (latitude and longitude) and projected coordinates (like UTM).
Reference Data: An accurate map or spatial dataset with known coordinates used as a base for alignment.
Essential Components
Control Points: These are the most critical components. They are specific, identifiable locations that exist in both the unreferenced image and a georeferenced dataset. They act as "anchors" to stretch and align the image correctly.
Qualities of a good control point: Easily identifiable, stable, and widely distributed across the image. Avoid points that might have changed over time.
Map Projection: A mathematical formula that translates locations from a 3D Earth to a 2D map. You must define both a coordinate system and a map projection for your georeferenced data to align correctly.
Transformation: The mathematical process used to convert the coordinates of the original data to match the coordinate system of the reference data.
Control Points: These are the most critical components. They are specific, identifiable locations that exist in both the unreferenced image and a georeferenced dataset. They act as "anchors" to stretch and align the image correctly.
Qualities of a good control point: Easily identifiable, stable, and widely distributed across the image. Avoid points that might have changed over time.
Map Projection: A mathematical formula that translates locations from a 3D Earth to a 2D map. You must define both a coordinate system and a map projection for your georeferenced data to align correctly.
Transformation: The mathematical process used to convert the coordinates of the original data to match the coordinate system of the reference data.
The Georeferencing Process
Basic Steps
- Add the unreferenced raster image to a GIS.
- Add a georeferenced dataset to use as a reference.
- Create control points by clicking a location on the unreferenced image and then the corresponding location on the referenced data.
- Choose a transformation method (e.g., affine, polynomial).
- Update the georeferencing to apply the transformation.
- Save the georeferenced image.
- Add the unreferenced raster image to a GIS.
- Add a georeferenced dataset to use as a reference.
- Create control points by clicking a location on the unreferenced image and then the corresponding location on the referenced data.
- Choose a transformation method (e.g., affine, polynomial).
- Update the georeferencing to apply the transformation.
- Save the georeferenced image.
Transformation Methods
Affine: Uses a minimum of three control points for linear scaling, shearing, and rotation. It is the most common method and works well for images with minimal distortion.
Polynomial: Uses more than three control points to account for complex distortions (e.g., curved maps). The higher the order, the more complex the transformation and the more control points are needed.
Spline: A "rubber-sheeting" method that locally warps the image to match the control points. It is excellent for historical maps with significant local distortions but can cause warping in areas without control points.
Affine: Uses a minimum of three control points for linear scaling, shearing, and rotation. It is the most common method and works well for images with minimal distortion.
Polynomial: Uses more than three control points to account for complex distortions (e.g., curved maps). The higher the order, the more complex the transformation and the more control points are needed.
Spline: A "rubber-sheeting" method that locally warps the image to match the control points. It is excellent for historical maps with significant local distortions but can cause warping in areas without control points.
Step-by-Step Guide on How Georeferencing is Performed
- Collecting Reference Data:
- Obtain the image or map that needs to be georeferenced.
- Gather reference data (coordinates of known locations) from reliable sources such as GPS devices, maps, or existing GIS data.
- Preparing the Image:
- Load the image or map into the georeferencing software.
- Ensure the image is properly oriented and cropped if necessary.
- Adding Control Points:
- Identify and mark control points on the image. These are points with known geographic coordinates.
- Control points should be well-distributed across the image to ensure accuracy.
- Entering Coordinates:
- Input the real-world coordinates for each control point.
- Use a consistent coordinate system and datum for all control points.
- Transformation:
- Apply a transformation method to align the image with the real-world coordinates. Common methods include:
- Linear (affine) transformation: used for simple adjustments.
- Polynomial transformation: used for more complex distortions.
- Projective transformation: used for images taken from an angle.
- Checking Accuracy:
- Assess the accuracy of the georeferencing by checking the residuals (differences between the actual and transformed coordinates of the control points).
- Adjust the control points or add more if necessary to improve accuracy.
- Saving the Georeferenced Image:
- Once satisfied with the accuracy, save the georeferenced image in the desired format.
- Export the associated geographic data for use in GIS applications.
- Collecting Reference Data:
- Obtain the image or map that needs to be georeferenced.
- Gather reference data (coordinates of known locations) from reliable sources such as GPS devices, maps, or existing GIS data.
- Preparing the Image:
- Load the image or map into the georeferencing software.
- Ensure the image is properly oriented and cropped if necessary.
- Adding Control Points:
- Identify and mark control points on the image. These are points with known geographic coordinates.
- Control points should be well-distributed across the image to ensure accuracy.
- Entering Coordinates:
- Input the real-world coordinates for each control point.
- Use a consistent coordinate system and datum for all control points.
- Transformation:
- Apply a transformation method to align the image with the real-world coordinates. Common methods include:
- Linear (affine) transformation: used for simple adjustments.
- Polynomial transformation: used for more complex distortions.
- Projective transformation: used for images taken from an angle.
- Apply a transformation method to align the image with the real-world coordinates. Common methods include:
- Checking Accuracy:
- Assess the accuracy of the georeferencing by checking the residuals (differences between the actual and transformed coordinates of the control points).
- Adjust the control points or add more if necessary to improve accuracy.
- Saving the Georeferenced Image:
- Once satisfied with the accuracy, save the georeferenced image in the desired format.
- Export the associated geographic data for use in GIS applications.
Assessing Accuracy and Troubleshooting
Root Mean Square Error (RMSE): This is the most common metric for assessing accuracy. It is a single value that represents the difference between the location of the control points you set and their new, transformed location. A lower RMSE indicates higher accuracy.
Number and Distribution of Control Points: Accuracy improves with more points and an even distribution. Points that are clustered in one area can lead to inaccuracies.
Common Challenges:
Old or distorted maps: Historical maps often have significant warping, tearing, or inconsistent scale.
Lack of clear control points: Some maps may lack identifiable features, making it difficult to find matching locations in the reference data.
Different projections or datums: The reference data and the raster image might use different projections or datums, which must be resolved for correct alignment.
Root Mean Square Error (RMSE): This is the most common metric for assessing accuracy. It is a single value that represents the difference between the location of the control points you set and their new, transformed location. A lower RMSE indicates higher accuracy.
Number and Distribution of Control Points: Accuracy improves with more points and an even distribution. Points that are clustered in one area can lead to inaccuracies.
Common Challenges:
Old or distorted maps: Historical maps often have significant warping, tearing, or inconsistent scale.
Lack of clear control points: Some maps may lack identifiable features, making it difficult to find matching locations in the reference data.
Different projections or datums: The reference data and the raster image might use different projections or datums, which must be resolved for correct alignment.
Real-World Applications
Georeferencing is essential for a variety of tasks:
Urban Planning: Georeferencing historical maps allows planners to track land-use evolution.
Environmental Studies: Researchers can georeference historical aerial photos to analyze changes in deforestation or glacial retreat.
Emergency Management: First responders can quickly georeference a scanned floor plan to aid in navigation and situational awareness.
Geology: Geologists can georeference old geological survey maps to integrate them with modern satellite imagery.
References- https://www.esri.com/about/newsroom/arcuser/understanding-raster-georeferencing
- https://desktop.arcgis.com/en/arcmap/latest/extensions/maritime-charting-guide/edit-dnc-enterprise/changing-the-geographic-coords-of-a-raster.htm
Georeferencing is essential for a variety of tasks:
Urban Planning: Georeferencing historical maps allows planners to track land-use evolution.
Environmental Studies: Researchers can georeference historical aerial photos to analyze changes in deforestation or glacial retreat.
Emergency Management: First responders can quickly georeference a scanned floor plan to aid in navigation and situational awareness.
Geology: Geologists can georeference old geological survey maps to integrate them with modern satellite imagery.
- https://www.esri.com/about/newsroom/arcuser/understanding-raster-georeferencing
- https://desktop.arcgis.com/en/arcmap/latest/extensions/maritime-charting-guide/edit-dnc-enterprise/changing-the-geographic-coords-of-a-raster.htm
Comments
Post a Comment