The over arching idea is to achieve being able to look at the views from the station in any direction, nadir, limb-ward and star-ward, as the views are available. The goal is to create a mechanism to allow organizing/tagging/crowdsourcing of information that would improve the usability of the imagery for science and educational projects. There is a sample set of images for use in the project development.
Detailed Description of Specific Subchallenges:
• Earth Views Georeferenced - The Earth views need to be georeferenced; there are a couple ways of doing this. 1. People can take an ISS image and a map or georeferenced satellite image and pick a point on the ISS image, then choose the corresponding point on the map or satellite view to assign that lat long to the ISS image. About 6 to 10 of these points would allow for the image to be stretched to a grid coordinate projection such as Google Earth or in image processing software used for GIS and comparing/overlaying imagery and other data. We can compute the nadir point of the ISS at the time the image was taken and provide that as some initial data, in addition to the lens size and approximate field of view (if the image is taken toward nadir).
• Crowdmapping - if people are traveling to a place on an ISS image, they could select a point on the image of where they are and then use their source of GPS to mark that spot on the image, similar to, or maybe in conjunction with, OpenStreetMap. Over time the GPS input could be compared and when the entries start converging they would be the more accurate GPS for those points. This would work better as the image is stretched over a known grid coordinate layer then other GPS positions not exactly on formerly entered ones could be verified on the accuracy.
• Earth Limb views referenced to star fields - Reference the ISS images with stars in the background to the associated star field and display them on a star field chart as if looking up in a planetarium. Over time we should get the area above the limb filled in for the different times of night and seasons. We have a good start now.
• Earth Views displayed in chronological perspective - this would work better if the images were georeferenced (stretched to a map) first which make them easier to compare quantitatively. But often you don’t need things lined up perfectly to notice that there is change. They would need at least a defined center point of the image to identify images which are in the same area. For this challenge we can provide ISS images with identified center points and camera file date and time taken.
