Challenge 18-001: Misalignment limiting automated building detection.


The objective of this challenge is to develop an algorithm that fixes alignment issues between building locations in Open Street Map and their location in Bing satellite imagery. 

Background info 

Missing Maps ( is an open, collaborative project in which volunteers help to map areas where humanitarian organizations are trying to meet the needs of vulnerable people by mapping buildings for example to improve and quicken humanitarian help during disasters . Various satellite imagery sources are being used by volunteers of the Missing Maps project to perform mapping tasks; most often Bing Maps or DigitalGlobe satellite imagery are used. Because of differences in geo-referencing between imagery sources, the Open Street Map (OSM) location of buildings mapped with DigitalGlobe will be misaligned when plotted on Bing Maps imagery and vice versa. The extent of the misalignment typically ranges from less than a meter up to dozens of meters. This misalignment issue currently prevents the 510 data team from the Netherlands Red Cross from using Open Street Map building data as ground truth for remote sensing techniques. If it were to be solved, we could use our Open Street Map data in combination with image recognition techniques to improve the capabilities of the Missing Maps project. Help us unlock the potential of the Open Street Map data! 

[wordpress_file_upload uploadpath=”uploads/challenge18-001/” uploadpatterns=”*.csv*” createpath=”true” placements=”title/filename+selectbutton+uploadbutton/message” uploadtitle=”Upload your result file” personaldatatypes=”userdata and files”]