Meet the new service – #Mapflow Data #API.

One of the major upcoming updates is the user's data management which we are implementing step by step into the Mapflow services.

Here is the first release of Mapflow Data API which is substantially extending the data management capabilities.

Now it's possible to create mosaics. What is "mosaic"? This is the collection of single images, aimed at instant preview and processing. It's helpful in such cases as mosaicing the number of single aerial images covering some area or batching the large orthomap into separate chunks for faster uploading and optimizing its storage in the cloud. The resulting mosaic layer is provided as COG available for XYZ/TMS tile requests.

More capabilities are related to image management: users can reuse their images for the next processings without repeating uploads, linking to the mosaics, adding metadata by applying key:value tags.

Mapflow Data API is absolutely free to use.

Contact us if you have questions or issues.

Just beautiful

Following the release of Mapflow Data API we keep on posting about new methods which are supposed to fulfill the next powerful integration of Earth Observation data sources into #Mapflow.
We added the initial scope of methods for #Imagery #Search #API. Getting started with this API, developers can search for available commercial satellite imagery by AOI coordinates, applying such filters as the date ranges and cloud cover and query the metadata by the Image ID. To check more and see the examples, visit API documentation.

To github users - Mapflow-QGIS 1.7.0 is already available in original repository. Soon it will be published for all QGIS users.
Join the project on github, add it to your starred repos if you like it, and don't hesitate to write us an issue.

Mapflow QGIS plugin v1.7.0 is released and published in official plugins repository!

The changes are not so dramatic, but it will make your experience more fulfilling as we have fixed some annoying bugs.
The most visible update is the data providers tab: it was reworked for better management of your own tile-based links, including your own #Maxar #SecureWatch subscriptions.
👉 Relaunch your QGIS to update the plugin.

#GeoQuiz No.8
#satelliteimagery #urbanforest
🌍 For real travelers!

How many cities can you recognize?

Send us your answers to [email protected]. As usual, we are ready to award you with free #Mapflow credits for research activities.

How you can get started with the open data by #Maxar using #QGIS.
This post was written in the context of the devastating #earthquake in Turkey and Syria.

We got the new urban study powered by the #Mapflow #Forest vegetation detection technology.
Andrey Pirogov - a dedicated geographer - studied 5 cities in Central Asia: #Almaty, #Bishkek, #Tashkent, #Ashgabad, and #Dushanbe - which one is the greenest one approved by #EarthObservation and #AI-#mapping? 🍃
Note that only imagery basemaps were used and vegetation was not perfect in all case studies.
This is also important that all data samples were published in our open data repository on Github.

We'd like to support the volunteer mapping during the #Turkey-#Syria #earthquakes aftermaths and recovering efforts.
❗️To this end we published all the images from #Maxar Open data to make them available to distribute from our tile server. As it seems to be the gap between the mapping tasks and the integration of satellite imagery updates into the mapping applications or GIS.
The difference is that we distribute these images by scenes (not by chunks) using our XYZ tile server, which can be connected to any GIS, #RapID #Openstreetmap, #Mapflow-#QGIS, whatsoever. The scenes are collected by Image IDs - this is the common and interoperable way to distribute satellite images.
👉 Find the list of all URLs to our server (and to the files on AWS, published by Maxar).
❗️ If you are helping in volunteer mapping and want to apply Mapflow AI to detect buildings in the affected areas - please reach out to us at [email protected]

One of the most impressive satellite pictures of the affected areas - the Earth crack in Olucak, Nurdağı, Southeastern Anatolia Region, Turkey (37.17486073, 36.71354923)

You can connect to open Maxar images using the corresponding URL from our tile server.

Find and preview this image in Openaerialmap

🌲 The newly updated Forest detection model is now available for all #Mapflow users!

We fine-tuned the model to provide more precise boundaries and detailed objects in specific areas, like mountains, arid zones, open woodland, etc. Some bugs in the previous model were fixed.

Note that it’s important to use the relevant by season #satellite imagery to obtain the most accurate results. With Mapflow you have instant access to the leading imagery providers and can easily find the most appropriate images.

Thanks to all users who leave their feedback during the beta testing of the updates, we do appreciate it! 🤝

For your inspiration - some studies that have been conducted during the testing of the updated model - “The greenest city centers in Central Asia” ( 👉https://medium.com/geoalert-platform-urban-monitoring/the-greenest-city-centers-in-central-asia-e174c4b06e82).

We are ready to grant additional credits to researchers who use our model for forest preservation projects. Please, contact us for more info.

Testing #Jilin satellite mosaic (0.75m, RGB) with the new #Mapflow #Forest model - in urban environment

#Tukey_erthquakes - added data layers to a single #QGIS project
https://github.com/Geoalert/Emergency_mapping

Our topic was selected for the OpenTalks.ai conference. This year the conference takes place in Yerevan, March 6-7.
Geoalert CTO Alexey Trekin will present on "Case-driven CV in satellite image processing"
> Despite being proven in the image processing field, neural networks still are a tricky tool for cartography. Can we trust the results? What should we do with the errors? Should we rely on selling ready models as a service, or stick to on-demand development?
In this talk I will share some practical cases: how do we derive the model for the particular task and area from the general off-the-shelf model, how do we collaborate with human cartographers and how to handle user's feedback.

We are thrilled to meet old friends and new people in the community. See you in Yerevan!

From the user #feedback.
#Forest model 🌲 in Montenegro 🇲🇪

Green city patterns in Europe
By using Mapbox Satellite basemap and #Mapflow AI, 45 urban green patterns were generated.
From Reykjavik to Valletta, from Dublin to Kyiv, the tree cover for each European capital reflects a unique mix of human-nature interaction.
Download the hi-res PDF from our public repo

Eugene joined our RnD team as a CV Engineer. 👏

We love Uzbekistan for its people and trees so we decided to support projects for social impact.
To this end we launched call to projects for Uzbekistan https://geoalert.io/projects
We are looking forward to supporting developers – how to create maps, tell stories with maps and sometimes even get insights about what is going around in the urban and natural environment.
Within this initiative, we've already started publishing open data and working with teams.
The initiative is supported by our friends from Gisgeo.org where you can find the description in Russian.

Turkish independent researchers published an article about the ways they tried to determine the age of the buildings in the areas affected by devastating #earthquakes and how many of them collapsed. As they didn't get any related information from official institutions, they tried remote sources such as satellite imagery, open data and the imagery search and analysis with #Mapflow and described their path. (The article is published in Turkish, use a translator).

> For this, it is an important method to apply the building detection model offered by MapFlow to the satellite images before and after the earthquake and to detect the difference with a geographic information systems software such as QGIS.

> In order to detect the situation before and after the earthquake with the help of artificial intelligence in a meaningful way, we had to have high resolution satellite images. We tried to get the current images via Maxar SecureWatch, where MapFlow works in an integrated way, and we obtained the images from Skywatch, a commercial organization that provides satellite imagery at the point where satellite images are missing.
However, we encountered some difficulties in practice. We have determined that the GeoTIFF files, which are the format of satellite images that are publicly shared from the earthquake-affected area and can be used for analysis, are not of sufficient quality.

We visited Yerevan for the AI & business conference. It's important to express our gratitude to the organizers for inviting us. Originally it's the largest of the Russian conferences dedicated to the advancements in AI in science as well as the implementation of it into the real business application and the most democratic one. This year the event's venue moved to Armenia, the conference got fewer participants but the same big corporations from Russia, but the spirit of the no-boundary science and digital community is still well spotted there.

BTW we met with our new teammates who came from Tbilisi.