SchoolTube Community
in the vast realm of earth observation, quantitative analysis plays a crucial role in understanding our planet's complex systems. today, we're diving into the world of segmentation algorithms and how they can help us make sense of the terrestrial nitrogen storage (tns) and other environmental phenomena. let's explore how the 2018 earth engine user summit (eeus) in dublin brought together experts to share insights and develop innovative solutions for analyzing earth's data.
segmentation algorithms: a non-traditional approach
at the eeus 2018, participants took a unique approach to the hackathon, focusing less on coding and more on image analysis. they prepared a code implementing new segmentation algorithms and invited users to work together, examining various images to determine the best parameters for each purpose. this collaborative effort aimed to help everyone, from beginners to advanced users, learn something new about earth engine and crowdsource algorithm parameters.
results and deliverables
the hackathon's efforts yielded some impressive results. participants were able to achieve satisfactory outcomes in different settings and on various image platforms. if you missed the event, don't worry! you can still get involved by contacting the organizers. they have a module to share that allows you to input your own parameters, images, and regions.
one of the hackathon's deliverables was to start scoping out the global forest change from the 1970s. this involved segmenting multispectral (ms) images using principal component analysis and keeping two bands. the segmentation looked great, and the team is now looking forward to segmenting all the images in earth engine.
tools and resources
the segmentation script is available in a module, which you can access by clicking on the provided link. the organizers will also share it on the summit group. additionally, they have developed sliders to collect more data and feedback on which parameters work best for specific purposes. this tool is a valuable resource for developers and experts alike.
comparing segmentation and classification
the hackathon participants compared the results of unsupervised classification with per-pixel classification on an mss image. while the comparison was qualitative, the users' feedback was crucial in determining the best approach. the goal is to gather more insights from experts and improve the segmentation algorithms further.
conclusion
quantitative analysis, particularly segmentation algorithms, is a powerful tool for understanding earth's complex systems. the eeus 2018 hackathon demonstrated the potential of collaboration and crowdsourcing in refining these algorithms and improving our understanding of the tns and other environmental phenomena. by working together and sharing knowledge, we can make significant strides in earth observation and contribute to a more sustainable future.
You may also like
https://www.schooltube.com/?p=34485