Data Capture – Georeferencing images and Digitising 

The aim of this project was to facilitate the widening and deepening of knowledge and understanding of coordinate systems and geodetic reference systems, spatial referencing, vector data structure and the concept of topology. 

Tasks:

• Georeference the “paper” soil map of Densu in southeast Ghana. 

• Digitise a roads and a soils vector layer from the paper map. Into Ghana metre grid coordinate system. 

• Present the digitised results in a cartographically sound manner. 

Methods:

Georeferencing method: 1st Order Polynomial (Affine) transformation applied.

Ground Control Points (GCPs): 11 points used.

• Reason for Affine choice:
  Less computationally intensive than higher-order transformations.
  Provides reliable accuracy for many map types (Longley et al., 2015).
  Preserves the original raster shape.
  
  

• Use of multiple GCPs:
  More than three GCPs included to improve transformation accuracy.
  Although mathematical error increased with more GCPs, overall accuracy improved.

Results

Forward Inverse Error Total RMSE Number of GCPs

403.323955 Metres 12.527959 Pixels 403.5183 Metres 11

Conclusion

This project demonstrates the power of GIS techniques in converting historical maps or maps without a defined coordinate system into modern, analysable digital data. The georeferencing and digitization of the Densu basin soil map have transformed a scanned map into a digital resource.