Georeferencing, Editing and Creating Vector Data
Georeferencing
Raster data is obtained from many sources, such as satellite images, aerial cameras, and scanned maps. Unlike modern satellite images and aerial cameras that tend to have relatively accurate location information and might need only slight adjustments to line up all your GIS data, scanned maps and historical data usually do not contain any spatial reference information. In these cases you need to use the process of georeferencing.
Georeferencing is the process of assigning real-world geographic coordinates to a raster image or a scanned map, enabling it to be accurately placed within a spatial reference system. This process involves matching identifiable points on the image with corresponding locations on a reference dataset, such as a satellite image or a vector map. For control points you should select features that are stable and easy to identify in both datasets such as churches, bridges, road intersections, river confluences, or long-standing public buildings. The points should be spread across the whole map sheet rather than concentrated in one corner.
Georeferencing is essential in cartography and GIS, as it allows historical maps, aerial photographs, or other spatial data to be integrated with other GIS layers for analysis, visualization, and decision-making.
Sources:
Vectorization
For analysing raster maps, it is almost always necessary to vectorise them, i.e., convert the map into vector data. There are various ways to automate this process, but here we will demonstrate the simplest method: manual vectorisation.
1. Vector data editing tools are located in the Edit tab.
2. New elements are created by clicking the Create and then selecting a drawing tool for the desired subtype in the Create Features pane.
3. Vectorized points are added with the left mouse button. To complete the vectorization of a particular element, either double-click with the left mouse button or select the icon Finish in the tools at the bottom of the screen. When vectorising, make sure snapping Snapping is enabled.
4. After editing vector data, do not forget to save your edits by clicking Save in Edit tab.
Sources:
Checking vector topology
If you want to check the topological integrity of vector data, all layers you want to validate must be stored within a single dataset.
1. To create a new topology, right-click the dataset → New → Topology.
2. On the first page of the Create Topology Wizard, define the topology parameters: its name, cluster tolerance/precision, and input layers.
3. The second page contains definitions of the topology rules to be checked. Set these as needed. In this example, we will check the rules Must Not Have Gaps (Area) (the data must not contain gaps), Must Not Overlap With (Area-Area) (layers must not overlap one another), and Must Not Overlap (Area) (features within a layer must not overlap themselves).
4. The third page provides a summary of the topology. Click Finish to create it.
5. If the topology does not appear in the output dataset, refresh its contents by right-clicking → Refresh.
6. Next, you need to validate the topology by right-clicking the topology in the Catalog pane → Validate.
7. After validation, drag the topology layer into the map view—you should see any detected errors.
8. Use the Edit tools to fix the highlighted errors in the original data. After editing, validate the topology again. If no errors are found, the checked layers are topologically correct.
Tips for speeding up topology checks:
- If possible, for data with the same attribute (e.g., forest, meadow) you can run Dissolve. This merges features and can remove self-overlap issues within the same category (for example, two meadow polygons overlapping each other).
- To simplify topology checking, you can merge all checked data into a single layer and then validate that layer on its own. This way, you do not need to check overlaps between different layers using Must Not Overlap With{: .outlined_code}; instead, you only check the new layer against itself using Must Not Overlap{: .outlined_code}. However, after checking, do not forget to fix any errors in the original data.
Sources:
Assignment 03
Vectorization of an old map
TASK:
Create a simple map reconstructing part of Prague in the early 20th century. In the map, distinguish at least four types of features: water bodies, green areas, built-up areas, and public spaces/streets. The map must include labels for at least one feature from each category, with label styling adapted to the feature type.
DATA SOURCES:
- Where can I find more old maps?
SUBMISSION FORM:
- 1 map in PDF format (submit by 29/03, send to petra.justova@fsv.cvut.cz)
INSTRUCTIONS:
Step 1: Georeference the map
- Create new project in ArcGIS Pro (save to disk :H).
- Add the old map to your Map project (add data).
- Find the added map (zoom to layer).
- Activate the Georeference tool (Imagery tab -> Georeference).
- On the Georeference tab, click Add Control Points. Now try to find at least 4 identical points (control points) on the image (source) and the reference map (target). These points should be spread out throughout the image to obtain the best possible registration (For example churches, old bridges, islands, towers...).
- After collecting all points, on the Georeference tab, click Save and Close georeference.
Step 2: Create new data
- Create new dataset (Catalog-Geodatabase-New-Dataset).
- Create new feature class (Catalog-Geodatabase-New-Feature Class). Create two polygon layers: (1) The first layer, named “extent”, will define the boundary of the area of interest. The second layer, named “LandUse”, will contain the vectorised polygons representing land use types.
- Create subtypes for “LandUse” layer (Attribut table-Table-Add-Subtypes-Create). Define following subtypes: Important building, Building, Public space/Street, Green area, Water body and Other.
- Set the symbology of “LandUse” layer by subtype (Save-Symbology-Unique values) and assign colours.
- Vectorize the boundary of the area of interest into “extent” layer.
- Vectorize all types of land use in your study area into a “LandUse” layer using simple editing tools.
- Merge the same subtypes into one element (Tools-Dissolve).
- Do not forget to check the topology. Fix any errors if necessary.
- Create a new annotation layer. In this layer, define four annotation classes that use different font styles and colours to distinguish labels for the four basic land-use categories, for example: buildings (bold, black/dark grey), public spaces/streets (regular, black/dark grey), green areas (regular, dark green), and water bodies (italic, dark blue).
- Label at least one feature from each category.
Step 3: Create a layout
- Create new Layout (A4 Landscape).
- Title.
- Map frame (in scale 1:5,000).
- Information about scale.
- Information about author.
- Legend - add with (Add legend-Convert to graphics-Ungroup) and edit.
- Try to add some labels to important places.
- Export Layout in PDF Format, DPI 120.
