With this webapplication you can collect trainingdata in an efficient way resulting in a machine learning process. The datacollecting is supported by a Layer which shows the AOA (area of applicability) provided by the AOA-algorithm in Hanna Meyers CAST-Package.
Clone this repository or copy it to a folder.
Then you need some basedata for your project. This thesis uses two airial Sentinel2 pictures downloaded from the * EartzhExplorer. To follow the thesis, download the area of Münster with these two pictures:
- L1C_T32ULC_A025316_20200427T104347
- L1C_T32UMC_A025316_20200427T104347
Once you downloaded them, you can start the preparation.r file at /src/r-script. With help of this you can prepare these two images for the application resulting in a rasterfile perfectly to start your first classification. When you are done preparing your projectarea, start the application:
To make sure everything is running fine, install following software on your operating system. You will find the used version of each tool. Feel free to use other versions but there is no guarantee that no further problems will arise. What you need:
- R-Studio and the used packages in the scripts at /src/r-script
- nodeJS (v12.16.3)
- mongodb (Version 4.2.0)
- python (Version 3.7.7) with
- GDAL (Version 3.0.2) installed with osgeo4w
To get the app running after cloning or downloading it, make sure, that all fileconnections are correct.
Then you have to prepare the R-Scripts at /src/r-script for usage. Check the fileconnections and make sure to link the directories to the directory of the repository. Make then sure to install all needed packages in the aoa.R file and the plumberAPI.R file. Then run the plumberAPI.R script and the plumber-API will start up. It is now ready to use and the actual app can run its functions.
Then change to the console and navigate to the repositorys folder. Then run:
npm install
and
npm run build
After buildung you are finally ready to start and enjoy the application:
npm start
Go to localhost:3000 in your favourite browser and see the resulting application.
While testing the application there where some issues with tiler running in the R-script. Therefor i had to set the tiler.options to find the needed files. If they are set as path as in the tiler documentation shown it should work but for me it didn't. So keep that in mind and watch out.
If you build a new project and you have access to the server-folder you can easiely change files as you wish. The current version of this app does not support investigation area maps, that got more than one band. If you would like to see a multiband rasterlayer on the leaflet map you have to insert the project tiles manually. For testing purpose, I used to generate tiles with QGIS and the 'Generate XYZ tiles'-tool. Here you have the option to take high resolutions and change the zoomlevels of the tiles. Mind that lower zoomlevels take much longer to generate because of the high amount of created tiles.
When generating is done you can just update the files of your chosen project in the dist/layers/PROJECTNAME folder.
Keep in mind that tiler generates TMS-tiles and the tilelayers have a parameter named tms that is set "true" to handle that. If you use 'Generate XYZ tiles' in QGIS check the TMS box to avoid possible mistakes.
If you got any problems with tiler or installation of any functionality to bring tiler to life, you can allways create tiles with QGIS and place them into the right folder.
To use the classification file at /src/r-script you have to convert the JSON-file to a shape-file first. Then you have to make sure that the polygons are repaired. Leaflet does not save clean polygons so what you have to do is to repaire them with a tool like "repaire geometries". After this the polygons are closed as they should. Furthermore you have to translate your classes to finit numbers. All you have to do is to translate each class into a number (e.g. 1, 2, 3 and 4) by putting them into the placeholderId column of your shape-file. That's all! Keep in mind to not take the class-column for classification but the placeholderId.
NOT YET IMPLEMENTED
To run the test:
npm test
...and many more
- Lukas Bäcker - Initial work - LukasBäcker
MIT License
Copyright (c) 2020 Lukas Bäcker
Permission is hereby granted, free of charge, to any person obtaining a copy of this software and associated documentation files (the "Software"), to deal in the Software without restriction, including without limitation the rights to use, copy, modify, merge, publish, distribute, sublicense, and/or sell copies of the Software, and to permit persons to whom the Software is furnished to do so, subject to the following conditions:
The above copyright notice and this permission notice shall be included in all copies or substantial portions of the Software.
THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.
- Christian Knoth
- Hanna Meyer (https://github.com/HannaMeyer)