There are again so many interesting things we would like to share with you. We don't want you to miss the opportunity to get an ESA-sponsored account and free access to high-resolution commercial satellite data. The Sentinel Hub Custom Script Contest ends today, so wrap up your scripts and submit them on time. Learn how to use Batch Processing to scale up your eo-learn workflow and create cloudless mosaics. Read new blog posts in our Area Monitoring series. Join our next webinar on Process API, and more. Not to miss any of it, sit back, grab a cup of coffee or tea 🙂 and scroll down for details.
Get a free Sentinel Hub account and access to commercial data!
As commercial datasets can be costly, requesting a Network of Resources (NoR) sponsorship is a great opportunity for researchers, as well as pre-commercial exploitation. The European Space Agency's research sponsorship program enables you to get an account and commercial data for free. You might benefit from having access to data with astonishingly high resolution. In addition, having a Sentinel Hub account will give you access to tools, such as the Configuration Utility and Batch Processing.
Send an application, describing your research and needs. The NoR team will evaluate your submission and, if you are successful, grant you access to free resources. To facilitate the process, we created a step-by-step submission guide for you.
In November we finalised the transition of EVALSCRIPT to V3. We were preparing for this move for more than seven months informing users about the change through various channels and preparing tools to make the whole process as smooth as possible. With more than quarter of a billion of various requests every month, we expected some issues. However, our efforts seem to be beneficial as most of the users successfully migrated their processes on time.
If you are not sure whether this change impacted you, we would kindly recommend you to check if the data you are getting from Sentinel Hub is still as expected. This is especially important for OGC requests if you are using transparency settings or are requesting 16-bit or 32-bit output.
Contact us via Forum if you need some help, we will be happy to assist.
The third round of the Contest ends today. So, wrap up your scripts and submit them before midnight.
The Contest introduces a new (optional) challenge. In addition to standard scripts, we are looking for examples based on a machine learning workflow developed using eo-learn. To learn more, visit our Contest page.
We were accepting early bird submissions until 20 September 2020 and announced the winner of the early bird award in October. The final winners will be announced on the official Contest web page by the end of December 2020.
How to generate harmonized ML features over huge areas and not become broke
Training machine-learning algorithms with satellite data requires a series of carefully designed pre-processing steps. The eo-learn Python package facilitates the whole process, letting you focus on data analysis. However, when working over large areas, the computing power needed often gets out of hand. To make data access easier and cheaper, we have tailored the eo-learn Python package to work with our Batch Processing API, allowing you to transfer your workflow to the cloud. Read our blog post to learn how to query and make your satellite imagery machine-learning ready without needing expensive computational resources, and for a fraction of the price of the regular Sentinel Hub API.
Perform preparatory analysis and use Sentinel Hub Batch Processing API within Requests Builder to ease your work!
If you find processing Cloudless Mosaics time-consuming or too complicated then check our blog post to see how to do it quickly with our Batch Processing API! Learn how to use the Requests Builder as an intuitive GUI for Sentinel Hub API requests to create, analyse and process your Batch Processing requests. Following an example from a regularly overcast area of the world, let us run you through our workflow, that involves the assessment of the available satellite data, a multi-temporal Batch Processing request and merging the tiles into a single Cloudless Mosaic.
Common Agricultural Policy Marker Service
Our blog posts on Area Monitoring are becoming more and more popular, and even more so with our latest submission of three new blog posts and one update! The first one shows our approach to detecting observation outliers, such as images with cloud shadows. With a very simple procedure, we were able to tag invalid observations and use this information to improve other markers. This improvement was also considered in the general update of the mowing detection algorithm.
The other two blog posts tackle the issue where parcel activity is not homogeneous, such as multiple crops growing on the same parcel, or localized mowing activity. The development of the homogeneity marker enables us to detect such parcels and treat them differently. For example, these are good candidates to test our new process of automatic parcel boundary detection, where an approach with deep-learning helps us draw the line between one growing crop and another.
Explore new EDC Notebooks, apps and data
Remember the Covid-19 Custom Script Contest? Check out the related Jupyter Notebooks on the Marketplace! For example, browse the Covid-19 Lockdown and Pollution (NO2), the AOT anomaly analysis during Covid-19 measures and the Contest winning Truck detection scripts.
There are no limits to what you can do! This Notebook demonstrates large scale machine learning, using batch processing and eo-learn. You can now access IACS data for several EU member states, as well as the newly added Copernicus Services data. If you're a developer, you'll be happy to know that the EDC backend services are now available for developers to use in their own applications, with the EDC Developer Sandbox app.
As part of the Euro Data Cube Covid-19 Custom Script Contest in April and May, we looked for ideas on how EO satellites can monitor the effects of the pandemic. As you might know, some of the winning ideas were invited to the RACE Upscaling Competition, getting an opportunity to be released to the public via the ‘Rapid Action on Covid-19 with Earth Observation dashboard’ – a joint initiative from ESA and the European Commission. Learn more about the best ideas which are being integrated into the RACE dashboard. Read our blog post.
Learn about different topics related to Sentinel Hub first hand
We introduced our first live Sentinel Hub webinar in September starting with Custom Scripts. Following the same format, we organised another one in October presenting OGC API with QGIS integration. We are happy to invite you to the third webinar - Sentinel Hub Process API - which will be held this Thursday, 3 December at 14.00 CET (13.00 UTC). See the agenda here and register for the webinar today.
Mark the dates for our presentations at the following events:
We are continuing to work remotely and mainly meeting all of you virtually. Don't hesitate to send us your feedback via Forum or social media. We look forward to hearing from you.
The Sentinel Hub team