What difference it makes after tuning the model for the specific territory.
By "tuning" we mean the additional training of the model using the relatively small number of labels. In this case we tuned the model for Saudi Arabia 🇸🇦. #buildings #saudi #segmentation

❗️Service downtime announcement

We are planning some maintenance on our servers, so all our services, including Mapflow.ai, Mapflow API and QGIS plugin will be unavailable on Saturday, October 22, for the 2-hour period in 15:00 - 19:00 UTC.

In the meantime, we are preparing some major releases that we will announce next week, stay tuned! 😉

One more picture from the Space and from our team.
This week as you may know started Diwali - one of the most beautiful and bright World fests. For us it also represents the victory of light over darkness, the life - over death and war if you wish.
Sometimes it's getting darker in the other parts of the World, but the light will overcome.

The #benchmarking project is going on.
We are testing buildings model against the manually labeled aerial images in #Africa, Zimbabwe. #Openaerialmap

We've updated the #pricing page of #Mapflow. Hopefully it's becoming more clear. (However the integrated billing system for all our application is still under development).
Pay attention that we added special terms for Academic license. Feel free to ask your questions we are ready to help.

Thanks to the collaboration with Andrey Pirogov (gisgeo.org) we started our new regular campaign ("#Mapflow weekly update" / https://lnkd.in/dbS8Ex35), based on the information from our social media and blog for the past week.
Subscribe to our #newsletter to stay tuned for the product updates and applications of #Machinelearning in #Geomatics.
https://mapflow.ai/subscribe_newsletter

#GeoQuiz
This photo was taken by Alex Hancharenka, CV engineer, originally from Minsk. Where it is? (City, place)
Correct answers as usual will be awarded with Mapflow credits 😊
Send it to [email protected] from your Mapflow account email

The correct answer to the previous #GeoQuiz (https://www.tg-me.com/geoalert/343) is Bukhara. The view from the water tower (designed by engineer Shukhov). The interesting notice about this that the sign reads the incorrect destination. According to the Google Earth it makes 3,704 km from Minsk to Bukhara by walk. Not to say that it makes less by flight (as the flights usually follow so called geodesic line - the shortest distance on the sphere / surface).
Stay tuned for the next GeoQuiz 😉

#frontend #react
❗️Geoalert (https://geoalert.io/team) is looking for a Frontend developer (Junior+ / Middle) to join our team in 📍 Tashkent.

Read more about openings: https://geoalert.io/docs/openings.pdf

Contacts to send your CV + github:
Email: [email protected]

Geoalert team is taking part in the mapping challenge.

Starting today we will be posting one map every day in accordance with the list of daily categories.

The theme of day 1 is "Points".

Here you see the heatmap based on Mapflow.ai free processings for the current year (the commercial accounts are excluded from this map due to NDA reasons). The map was developed in #QGIS using the conversion to centroids (still the original AOIs of the processings are presented as polygons) to visualise it as a heatmap. The rest is simple – the resulting heatmap is weighted by the cost of the processing which is relevant to its area. Cartodb is used as a basemap.

This is one of the ways you can analyse users aggregated data, especially when it comes to geospatial / mapping services.

Please tell us what you think and stay tuned for next maps and #30DayMapChallenge.

#Grid city patterns. This will be the map for Day 2 of #30DayMapChallenge.
The category is "Lines".
Actually, the place was suggested by one of the Mapflow users who shared the result of roads segmentation in his post (it's Balcarce, Argentina - seems that they do like grid structures).
The roads extraction in Mapflow is translated through the following stages:
- raster mask of roads in satellite image
- skeletonization of raster mask to get centerlines of the roads
- vectorization to get GeoJSON
With this Road model we got stuck in experimental mode - so don't take it very seriously, cartographers can do much better.
However, you can find city patterns and analyze their characteristics very fast 😉. E.g. we got the total length of the street network within the admin boundaries of Balcarce as 262 km.

❗️Due to maintenance on servers all our services, including Mapflow.ai, API, and Mapflow-QGIS will be unavailable for 2 hours in the period from Nov 4, 2022 9:00 UTC until Nov 4, 2022 20:00 UTC

Belated #30DayMapChallenge, Day 4: Green 🍏

This place shows the guitar-shaped forest in Argentina that the farmer manually planted in memory of his wife.
You can easily google the story.

We were interested in applying the #forest mask extraction #Mapflow model to this area. Here is the resulting map that we got.

Bless the farmer and his forestry music 💚

One more map for today's #30DayMapChallenge, Day 5: Map about #Ukraine. 🇺🇦

This is the area near Kharkiv (50.0705, 36.4820), switching between different stages of processing of #Sentinel-2 satellite image:
1) NDVI (colored with Ukrainian flag color palette)
2) TrueColor
3) Field mask extracted from the same Sentinel with enhanced resolution

All was done in #QGIS (using #Mapflow-for-QGIS plugin to download Sentinel and apply the field mask extraction). For the resolution enhancement we used #Skycues who provided us with the testing account.
Having said that the crop field segmentation model is still in the experimental mode and requires time series and manual validation for stable results, it can be combined with high resolution (or enhanced one) to achieve better accuracy.