The Map Accuracy Tools website was developed to facilitate the calculation of accuracy metrics related to the development of land cover maps from remote sensing. When land cover maps are validated, the ground truth data are compared with the values from the classified map, and a confusion matrix is then produced.
Users can upload their confusion matrix to the website and a range of different accuracy statistics will be calculated. More advanced users can download the R source code and run it locally. The tool can also be used in a teaching capacity to help students understand the concepts of map accuracy assessment.
A second motivation for the development of this tool was related to observations made by Salk et al. (2018) regarding common mistakes and pitfalls that occur in thematic map accuracy assessment.
The tool comes with two example confusion matrices for users to play around with, classes can be merged to see the effect (e.g., the deciduous and evergreen forest classes can be merged into a single forest class) or users can read the paper by Salk et al. (2018) and learn more about good and bad practices related to accuracy assessment. Such examples can also be incorporated into higher education curricula on remote sensing.
For users who want the complete validation workflow, i.e., uploading a land cover map, generating a statistically robust validation sample, interpreting the sample using very high-resolution imagery, and producing an accuracy assessment report, are referred to the LACO-Wiki tool for this additional functionality.
Last edited: 26 July 2021
Program Director and Principal Research Scholar Strategic Initiatives Program
Principal Research Scholar Novel Data Ecosystems for Sustainability Research Group - Advancing Systems Analysis Program
Senior Research Scholar Novel Data Ecosystems for Sustainability Research Group - Advancing Systems Analysis Program
International Institute for Applied Systems Analysis (IIASA)
Schlossplatz 1, A-2361 Laxenburg, Austria
Phone: (+43 2236) 807 0 Fax:(+43 2236) 71 313