Crowdsourcing EO datasets to improve cloud detection algorithms and land cover change

14857 28 79287 2 219037 document 79287 EO Open Science 2017 - Aleksandrov et al - Crowdsourcing clouds.pdf application/pdf 5d683ced33d2684e8a7e5eff4e47a6f3 MD5 28857 2017-09-29 10:10:36 https://pure.iiasa.ac.at/id/eprint/14857/1/EO%20Open%20Science%202017%20-%20Aleksandrov%20et%20al%20-%20Crowdsourcing%20clouds.pdf 14857 1 1 application/pdf text en public cc_by

EO Open Science 2017 - Aleksandrov et al - Crowdsourcing clouds.pdf

published 79288 2 219040 document 79288 Croudsourcing_EO_Datasets.pdf application/pdf 00ea296bd65cc01dd5b8068c06dd835e MD5 1860029 2017-09-29 10:11:39 https://pure.iiasa.ac.at/id/eprint/14857/2/Croudsourcing_EO_Datasets.pdf 14857 2 2 application/pdf slideshow en public cc_by

Croudsourcing_EO_Datasets.pdf

published 79289 2 219044 document 79289 lightbox.jpg image/png e8013bab56f1824b30715420018489e5 MD5 54513 2017-09-29 10:12:03 https://pure.iiasa.ac.at/id/eprint/14857/3/lightbox.jpg 14857 3 3 image/png other Thumbnails conversion from slideshow to thumbnail_lightbox en public

lightbox.jpg

http://eprints.org/relation/isVersionOf https://pure.iiasa.ac.at/id/document/79288 http://eprints.org/relation/isVolatileVersionOf https://pure.iiasa.ac.at/id/document/79288 http://eprints.org/relation/islightboxThumbnailVersionOf https://pure.iiasa.ac.at/id/document/79288 79290 2 219045 document 79290 preview.jpg image/png 52879da69423395d856db12bed01706e MD5 26071 2017-09-29 10:12:05 https://pure.iiasa.ac.at/id/eprint/14857/4/preview.jpg 14857 4 4 image/png other Thumbnails conversion from slideshow to thumbnail_preview en public

preview.jpg

http://eprints.org/relation/isVersionOf https://pure.iiasa.ac.at/id/document/79288 http://eprints.org/relation/isVolatileVersionOf https://pure.iiasa.ac.at/id/document/79288 http://eprints.org/relation/ispreviewThumbnailVersionOf https://pure.iiasa.ac.at/id/document/79288 79291 2 219046 document 79291 medium.jpg image/png 9af971c57f1042c0bc2578dc6d081132 MD5 9312 2017-09-29 10:12:05 https://pure.iiasa.ac.at/id/eprint/14857/5/medium.jpg 14857 5 5 image/png other Thumbnails conversion from slideshow to thumbnail_medium en public

medium.jpg

http://eprints.org/relation/isVersionOf https://pure.iiasa.ac.at/id/document/79288 http://eprints.org/relation/isVolatileVersionOf https://pure.iiasa.ac.at/id/document/79288 http://eprints.org/relation/ismediumThumbnailVersionOf https://pure.iiasa.ac.at/id/document/79288 79292 2 219047 document 79292 small.jpg image/png f4e679c2796cb6bbab27625ca92611b3 MD5 2206 2017-09-29 10:12:06 https://pure.iiasa.ac.at/id/eprint/14857/6/small.jpg 14857 6 6 image/png other Thumbnails conversion from slideshow to thumbnail_small en public

small.jpg

http://eprints.org/relation/isVersionOf https://pure.iiasa.ac.at/id/document/79288 http://eprints.org/relation/isVolatileVersionOf https://pure.iiasa.ac.at/id/document/79288 http://eprints.org/relation/issmallThumbnailVersionOf https://pure.iiasa.ac.at/id/document/79288 79293 1 219048 document 79293 indexcodes.txt text/x-c++ 22068e304315164ea766bc87b91b2ddb MD5 1564 2017-09-29 10:12:07 https://pure.iiasa.ac.at/id/eprint/14857/7/indexcodes.txt 14857 7 7 text/x-c++ other Generate index codes conversion from text to indexcodes en public

indexcodes.txt

http://eprints.org/relation/isVersionOf https://pure.iiasa.ac.at/id/document/79288 http://eprints.org/relation/isVolatileVersionOf https://pure.iiasa.ac.at/id/document/79288 http://eprints.org/relation/isIndexCodesVersionOf https://pure.iiasa.ac.at/id/document/79288 79305 2 219079 document 79305 lightbox.jpg image/png 772b41e84a5a045835083f0e63e2440a MD5 45125 2017-09-29 10:20:56 https://pure.iiasa.ac.at/id/eprint/14857/9/lightbox.jpg 14857 9 9 image/png other Thumbnails conversion from text to thumbnail_lightbox en public

lightbox.jpg

preview.jpg

medium.jpg

small.jpg

http://eprints.org/relation/isVersionOf https://pure.iiasa.ac.at/id/document/79287 http://eprints.org/relation/isVolatileVersionOf https://pure.iiasa.ac.at/id/document/79287 http://eprints.org/relation/issmallThumbnailVersionOf https://pure.iiasa.ac.at/id/document/79287 archive 353 disk0/00/01/48/57 2017-09-29 10:14:55 2025-03-12 05:00:25 2017-09-29 10:14:55 conference_item show 1 Aleksandrov M. Batic M. Milcinski G. See L. 8571 0000-0002-2665-7065 Perger C. 1899 Moorthy I. 2089 0000-0002-3664-3855 Fritz S. 8130 0000-0003-0420-8549 Crowdsourcing EO datasets to improve cloud detection algorithms and land cover change pub prog_esm Involving citizens in science is gaining considerable traction of late. With positive examples (e.g. Geo-Wiki, FotoQuest Austria), a number of projects are exploring the options to engage the public in contributing to scientific research, often by asking participants to collect some data or validate some results. The International Institute for Applied Systems Analysis (IIASA), with extensive experience in crowdsourcing and gamification, has joined Sinergise, Copernicus Masters 2016 winners, to engage the public in an initiative involving ESA’s Sentinel-2 satellite imagery. Sentinel-2 imagery offers high revisit times and sufficient resolution for land change detection applications. Unfortunately, simple (but fast) algorithms often fail due to many false-positives: changes in clouds are perceived as land changes. The ability to discriminate of cloudy pixels is thus crucial for any automatic or semi-automatic solutions that detect land change. A plethora of algorithms to distinguish clouds in Sentinel-2 data are available. However, there is a need for better data on where and when clouds occur to help improve these algorithms. To overcome this current gap in the data, we are engaging the public in this task. Using a number of tools, developed at IIASA, and Sentinel Hub services, which provide fast access to the entire global archive of Sentinel-2 data, the aim is to obtain a large data resource of curated cloud classifications. The resulting dataset will be published as open data and made available through Geopedia platform. The gamified process will start by asking users if there are clouds on a small image (e.g. 8x8 pixels at the highest Sentinel-2 resolution of 10 m/px), which will provide us with a screening process to pinpoint cloudy areas, employing Picture Pile crowdsourcing game from IIASA. The next step will involve a more detailed workflow, as users will get a slightly larger image (e.g. 64x64 pixels) and will then be asked to delineate different types of clouds: opaque clouds (nothing is seen through the clouds), thick clouds (where the surface is still discernible through the clouds), and thin clouds (where the surface is unequivocally covered by a cloud); the rest of the image will be implicitly cloud-free. The resulting data will be made available through the Geopedia portal, both for exploring and downloading. This paper will demonstrate this process and show some results from a crowdsourcing campaign. The approach will also allow us to collect other datasets in a rapid and efficient manner. For example, using a slightly modified configuration, a similar workflow could be used to obtain a manually curated land cover classification data set, which could be used as training data for machine learning algorithms. 2017-09 published http://eoopenscience.esa.int/ 276 228 211 98 public paper Earth Observation Open Science 2017 Conference Frascati, Italy 25-28 September 2017 conference FALSE FALSE no info:eu-repo/semantics/conferenceObject