We are happy to announce the latest addition to this KB Lab are now available: the dataset Classified Historical Newspaper Images (CHRONIC) and its counterpart CHRONReader.
CHRONIC is a dataset of metadata for 313.000 classified images harvested from Delpher’s Dutch digitised newspapers for the period 1860-1922. Thomas Smits and Willem-Jan Faber created the CHRONIC database by applying several Computer Vision techniques to classify the images. CHRONIC was originally created to test if state of the art computer vision techniques could be applied to historical images and research when Dutch newspapers started to use photographs instead of drawings to visually represent the news. CHRONIC is freely available via our Github.
CHRONReader enables you to find images in Delpher’s digitised newspaper in two new ways. First of all, based on the CHRONIC database of classified historical Dutch newspaper images, you can search for specific (combinations) of visual characteristics. First of all, you can indicate in which time period you want to search (1860-1922 available). Then you select if you are looking for drawings, photographs, or both of these categories. Thirdly, you can select one of the categories (any, buildings, cartoons, chess, crowds, logo’s, maps, schematics, sheet music, weather reports). Because all images in CHRONIC were classified for faces, you can also select if you want to search for images with a face.
CHRONIC and CHRONReader were developed in the researcher-in-residence project of Thomas Smits. For more information about the project, please see the blog posts he wrote while in residence at the KB Lab.