Annotated images of the Dutch coast
datasetposted on 27.08.2014 by Bas Hoonhout, M. (Max) Radermacher
Datasets usually provide raw data for analysis. This raw data often comes in spreadsheet form, but can be any collection of data, on which analysis can be performed.
188 images of the Dutch coast taken during the summer of 2013 at 4 different locations. The images are segmented into 600 superpixels using the SLIC algorithm. The superpixels are subsequently annotated in the classes sky, water (sea), water (pool), sand (beach), sand (dune), vegetation, object (sea), object (beach) and object (dune). The dataset can be used to train, validate and test coastal image classification algorithms like Logistic Regressors, Support Vector Machines or Conditional Random Fields.