LACO-Wiki is a new web-based solution for validating land cover and land use maps. Using a variety of reference layers including satellite and aerial imagery from Google and Bing as well as OpenStreetMap, validation is a simple four-step process. After uploading your dataset, generate and validate the samples and create a report with the accuracy assessment.
This latest version of LACO-Wiki was released in March 2020. Please use this version even though it may not contain any data sets, samples, validation sessions or accuracy reports that you uploaded or generated in the past. You can still access these files in the old version of LACO-Wiki at:https://old.laco-wiki.net. Note that we will migrate your data sets, samples, validation sessions and accuracy reports in the future. We will let you know when this migration has occurred. In the meantime, please upload any old data sets to this new version if you wish to continue working with them. If you have any questions, please email us at:[email protected].
Share your validated samples with us and you will help to build an open database that can be used to improve future land cover and land use maps.
From dataset ingestion to confusion matrix — the full validation pipeline in one open platform.
You can upload your own maps for validation in either vector or raster format. Currently accepted formats are shape files and geoTIFFs in a WGS84 projection. Once uploaded you can design a customized legend for display. Additional datasets can also be uploaded to help you in the validation process.
Once you have uploaded your map, you can create sets of validation samples using random, stratified or systematic sampling. You can specify the size of each sample or be guided by calculations of the minimum sample size needed based on the required confidence levels for your project.
Using reference information such as satellite imagery, you can validate your sample using your own legend, either by selecting the class, confirming the class or correcting incorrectly classified ones. You can validate the samples by yourself or you can share any validation session to distribute the work.
After validation you can download the raw data, the confusion matrix and generate a customized report on accuracy assessment, choosing from a set of different quality indicators including overall accuracy, omission and commission errors, kappa, average mutual information (AMI) and more.
Upload your own map or run a confusion-matrix analysis against an existing one — both routes are free for research.
Try out the new Map Accuracy tool that produces a confusion matrix and related metrics based upon your data.
Open calculator