QGIS (qgis.org) is the leading and most popular Open Source Desktop GIS. Users can visualize, manage, edit, analyse geodata, and compose printable maps. Get a first impression with a more detailed feature list.
Know more about QGIS and istall it from official site.
QGIS has an interface for external Python plugins that allows to connect more apps and extend core functionallity. Our “Mapflow - QGIS” app enables connection to Mapflow Mapflow processing API to run AI-mapping and add output as layers to the QGIS workspace.
Click Plugins -–> Manage…, go to the Not installed (or All) tab and enter “MapFlow” in the search box. Click Install Plugin. You will be able to see then if the newer version of the app is available (in the Upgradeable tab) and to check the changelog for details.
The plugin icon has appeared in the QGIS Toolbar.
If the icon isn’t automatically displayed, right-click on the Toolbar and check if the Mapflow checkbox is activated.
Main plugin workspace has two sections: left sidebar with the processing controls and the tabs section.
Processing controls panel allows to start new processing and/or rate finished processings, and includes following:
Processing controls panel
Name of the field / button
Description
Processing name
Name of your processing
Area
The area to be processed. This layer is automatically displayed in the drop-down list from the list of QGIS vector layers.
Data source
Base imagery to be processed. By default Mapbox is selected, in the drop-down list you can also select satellite images of the commercial providers. You can open your image through the additional options button.
Use image extent
This area becomes active if processing is performed on the .tif file. Allows you not to create an additional vector layer processing area for this image but takes its extent.
Processing area
Automatic calculation of the area of the selected processing area.
AI Model
Processing type. In the drop-down list, you can select the following processing types (default list of processing scenarios): Building Detection, Roads Detection, Forest Detection, Forest Detection With Heights, Construction Detection, Fields Detection.
Your balance
Remaining area available for processing
Start processing
New processing start button. You will see a notification about the successful start of processing, or about incorrectly selected parameters after clicking it. It will appear in the block for displaying and working with processing after a successful start of processing.
Stars for rating
Rating scores for the finished processing
Feedback field
Comments about rating
Your current balance is dispayed in the Tob bar. It also contains menu to access you personal profile on Mapflow.ai: top up you balance; open billing history; log out of current session.
SecureWatch is a service that provides global access to high-resolution satellite images and imagery basemaps from the world leader in remote sensing, MAXAR, through the subscription model. The spatial resolution of images varies in the range from 30 cm to 1 m. All images are accompanied by metadata, including information about the acquisition date and time, cloud cover etc. In our application we implemented the special interface to connect to this service and use imagery via Mapflow’s processings pipelines.
Use of embedded Maxar SecureWatch for image processing by Mapflow
In the Data source drop-down list, select the required Maxar product (SecureWatch, Vivid), switch to the Imagery search tab.
Maxar preview
Select your AOI in the Area drop-down list and click on the Search imagery.
Double click on the selected image in the search results (or click Preview button) to add it on the map.
Important
In the free tariff plan the Max zoom is limited up to 12 and the processing cannot be started using SecureWatch. If you want to use this data provider - you have to switch to the Premium tariff plan or write to us to get a quote.
Using your SecureWatch account for image processing by the Mapflow
Click + button and choose Maxar WMTS option in the dropdown list;
Enter Login / Password from your Maxar SecureWatch account;
Enter WMTS URL link for Maxar Secure Watch (SecureWatch - Login - Securewatch - Use with - Web Services - WMTS)
Optional: specify the coordinate system (default epsg:3857);
In the upper menu select Use With >> Web Services >> WMTS
Copy the WMTS (or TMS) url.
The Connect ID is different for each product you have in your SecureWatch subscription. Therefore, initially choose the one you want. To do this, open the User Profile menu and in the title bar select the required of the two suggested mosaics (Vivid and SecureWatch).
Click Preview.
Now the Maxar layer is available for preview in your raster layers list and for the AI-mapping processing using Mapflow.
How to find and process the image by Feature ID using Maxar SecureWatch
You can use SW to discover available images for you area of interest.
Go to the Providers tab.
Select Maxar SecureWatch from the dropdown list.
In the Maxar SecureWatch imagery Catalog select the vector layer containing the boundary of your area of interest.
Note
To define the imagery search area you can create the new polygon (Layer -> Create layer -> …, select Polygon as a geometry type, add polygon using the tool Add polygon feature) or upload it from the file with coordinates. If there is more than one polygon in the file, select with the tool Select object(s) the polygon you need. For more information on creating and working with vector layers, see the QGIS User Guide.
Alternatively, you can check the option “use canvas extent”.
Search imagery, to view meta-data of all available images intesecting your AOI. You can apply search filters and specify the period for which you would like to receive images. This will help in forming an imagery catalog with the necessary parameters.
Select the prteferable image from the imagery catalog or use the WFS generated vector layer (Maxar SW metadata) to search through more attributes. If you want to process a specific image in advance, insert your image ID in the field on the top of the plugin, this will make it easier to find the image in the imagery catalog.
Note
Imagery metadata is saved in the form of vector layer. You can interact with its Attribute Table by searching through all attributes.
Click Preview to view the selected image in the form of new raster layer (or double-clicking on the row in the table).
Attention
“max zoom 12” checkbox is active to prevent the paid streaming on the side of Maxar SecureWatch.
For example, let’s use the OpenAerialMap is an open collection of UAV imagery data, crowdsourced by users. The project is supported by a consortium of companies developing open source software and services for working with spatial data.
As soon as your aerial image is published on OpenAerialMap it’s presented on the public map and can be fetched using TMS/WMTS protocols.
Select the image and copy link to TMS.
Go to the plugin, on the Providers tab click on the Add (1) and enter the relevant data in the opened window (2). Click the Preview (3) the image, - you must be at the correct zoom and coordinates to see the image.
To start processing using this data source, go to the Processing tab, fill in all fields of processing parameters, click Start processing.
Name of the field / button
Description
Name
Name of other imagery data provider
URL
URL of the imagery data provider
Type
Data source type. You can enter your custom imagery source URL in one of the following formats: XYZ, TMS, WMS, Quadkey. All formats represent the most widely used protocols to fetch georeferenced imagery via HTTP.
You can upload your own GeoTIFF. All raster layers currently loaded in your QGIS (1) are visible in the drop-down list (2) and can be selected for upload. Opening files from the Additional options button (3) also adds it to the tree of QGIS layers.
Important
Please, follow the requirements specified on the page with Requirements for using the platform when uploading your own images for processing through the API of the Mapflow platform.
Check the data type
The Data type must be Byte (8 bit). If the Data type is Int16, or Float32 etc, please follow the instruction Userguides - How To.
Alternative option: use the preprocessing script
Check the image size
Both sides image dimension must not exceed 30.000x30.000 pixels. If you have larger images you should either cut them in smaller chunks or switch to the Mapflow custom.
Check the number of channels
Normally, the Mapflow processes 3-channel RGB rendered images. Mapflow platform can also process single-band (panchromatic) imagery, but the NN models are not
tuned for such kind of data, so the quality of the result may be worse than expected.
Check the projection and georeference
Make sure that your imagery is georeferenced in geographic or projected coordinate system.
Check the resolution
The resolution restrictions vary for different models, see Model requirements
You can send a request for data preprocessing to help@geoalert.io
The processing results that are 100% complete can be downloaded as a vector file to your local directory and automatically added as layer to QGIS workspace.
Double click on the processing name in the 1. Processing Table or select it and push the button “Download results”.
The layer will appear in the Layers panel (QGIS –> View –> Panels –> Layers) in the folder “Mapflow”.
You can work with it further as with the usual vector layer in QGIS.
Note
If the default AI model is used, the plugin automatically assigns predefined styles to the vector layer. For all custom models / pipelines the single default style is assigned. You can always change it.
To delete the processing - select it in the list, click the button “Delete” and confirm.
Warning
The processing cannot be restored by user. Before the permanent deletion the data backup is temporarily stored on the Mapflow server in case of emergency. So if you deleted your results by mistake and want us to restore the processing – send your request to the support without delay.
❗️ This option is available for the limited number of Mapflow customers who have signed up for the enterprise support
If activated, every completed processing comes with the status “Review required”. The user can either accept the result or request the review, so the support can reprocess it and get better results, until it meets the requirements.
To request the Review of the results:
Select the processing with the Status “Review required” and click the “Review” button
Provide comments in the Review dialogue:
Add your comment on why and what you want us to make a review
Optionally Add the polygon area, highlighting the objects that are not correctly processed, weren’t identified correctly, etc – it will help us to pay attention to the specific issues