Image Matching Web Interface Game
Deshpande, Prachi 1984-
MetadataShow full item record
Natural ecosystems are the backbone of human society. They support humanity's agricultural needs and provide clean air and water. However, environmental problems such as global warming and rising sea level created a strong urge to preserve natural ecosystem. The goal of the Virtual Prairie project is to study the fundamental mechanisms involved in regulating the population of plants in a prairie. The idea is to involve the general public in ecology projects, first training them in identifying prairie species and then allowing them to identify these species. The data generated are further processed to answer research questions such as how vegetation responds to variation in precipitation, effects of salt water on plant growth, etc. In an experiment conducted in Sapelo Island, Georgia, a large amount of data was recorded by digitally capturing images of a sampling area. The area was approximately 3,200 square meters and 3 to 4 images per square meters were acquired. Analyzing this large amount of spatially explicit digital data is impossible for a single scientist, creating a huge obstacle to progress of ecological studies. It would be helpful if the scientists were able to identify, for example, which plant species dominated over a few months, how it affected the animal community, what was the effect on their natural habitat, etc. Along with the detailed information that is available from the individual image, it is also advantageous to have a complete view of the entire marsh in order to understand the spatial relationships of different species. Therefore, stitching multiple overlapping images together to form a mosaic was necessary. However, applying automatic methods for image alignment and stitching did not produce accurate results because of the issues imposed by prairie images. Hence, the Image Matching Web Interface (IMWI) game was developed. IMWI is used as platform for users to be able to find matching points between pairs of images. We collect the data produced by the trained players playing on unprocessed images. These data will be used by a post-processing group to stitch the overlapping images together to form a mosaic of the marsh.