Ecosyste.ms: Awesome
An open API service indexing awesome lists of open source software.
https://github.com/dernasherbrezon/georef-lrpt
Georeference images received via LRPT
https://github.com/dernasherbrezon/georef-lrpt
lrpt maps openstreetmap satellite-imagery
Last synced: about 1 month ago
JSON representation
Georeference images received via LRPT
- Host: GitHub
- URL: https://github.com/dernasherbrezon/georef-lrpt
- Owner: dernasherbrezon
- License: apache-2.0
- Created: 2023-04-12T14:35:59.000Z (over 1 year ago)
- Default Branch: main
- Last Pushed: 2023-04-16T09:41:55.000Z (over 1 year ago)
- Last Synced: 2024-05-02T05:09:17.871Z (8 months ago)
- Topics: lrpt, maps, openstreetmap, satellite-imagery
- Language: Java
- Homepage:
- Size: 2.3 MB
- Stars: 0
- Watchers: 2
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
- Funding: .github/FUNDING.yml
- License: LICENSE
Awesome Lists containing this project
README
[](https://github.com/dernasherbrezon/georef-lrpt/actions/workflows/build.yml) [](https://sonarcloud.io/summary/new_code?id=dernasherbrezon_georef-lrpt)
# About
georef-lrpt application can georeference images acquired via [LRPT](https://en.wikipedia.org/wiki/Low-rate_picture_transmission) protocol. The result will be a .png file and a .vrt file. VRT file can be used by gdal to create GeoTIFF.
# Features
* Supported satellites: METEOR-M 2 (NORAD ID: 40069), METEOR-M 2-2 (NORAD ID: 44387)
* 3 channels + 1 alpha channel
* Alpha channel is generated for those regions where was no packets received for all 3 other channels
* Detect extents for manual split on anti-meridian
# How to use
## Step 1. Obtain LRPT binary files
These files should contain VCDU (data link layer frame) as they were received from the satellite. Find examples of such files in the ```src/test/resources/``` directory.
> Note:
> VCDU frames can carry variable number of image packets. So better to have a single binary file with all VCDU sorted in the ascending order or several files with filenames in the ascending order.
## Step 2. Obtain TLE
[TLE](https://en.wikipedia.org/wiki/Two-line_element_set) of the satellite. It should be in the 2-line or 3-line format.
> Note:
> TLE should be generated for the same day as binary files for better alignment precision.
## Step 3. Run
Get the help of all supported parameters using the command below:
```
java -jar georef-lrpt.jar --help
```
Example run:
```
java -jar georef-lrpt.jar --output-dir . --tle-file src/test/resources/2022-12-20.txt --vcdu-files "src/test/resources/*.vcdu"
```
The output will contain .png file and .vrt file.
## Step 4. Convert .vrt into GeoTIFF using gdal
Use [gdal](https://gdal.org) to convert .vrt into GeoTIFF:
```
gdalwarp -tps -overwrite -of GTIFF 2022_12_20_20_25_30_134_-14_162_-8.vrt 2022_12_20_20_25_30.tiff
```
The command above will create GeoTIFF file in the EPSG:4326 projection ( longlat WGS84 ). This projection can be further converted into epsg:3857 to use from online maps like [Google maps](http://maps.google.com/) and [OpenStreetMap](https://www.openstreetmap.org).
## Anti-meridian
GDAL and QGIS cannot detect if image is wrapped around projection. Thus special handling is needed.
1. Extract geodetic coordinates from the .vrt filename. georef-lrpt will generate .vrt filenames according to the format: ```____.vrt```. Please note that each coordinate can be negative. For example: ```2022_12_20_20_25_30_134_-14_162_-8.vrt``` Has longitude from 134 to 162 (west to east) and latitude from -14 to -8 (south to north).
2. Generate 2 separate tiff files from the same .vrt using gdalwrap. For the filename ```2022_12_20_18_44_50_142_-55_-173_-11.vrt``` two commands need to be executed ```gdalwarp -tps -overwrite -te 142 -55 180 -11 -of GTiff 2022_12_20_18_44_50_142_-55_-173_-11.vrt 2022_12_20_18_44_50-left.tiff``` and ```gdalwarp -tps -overwrite -te -180 -55 -173 -11 -of GTiff 2022_12_20_18_44_50_142_-55_-173_-11.vrt 2022_12_20_18_44_50-right.tiff```
Then left and right images can be added to the map. On the screenshot below the right image is drawn without alpha channel, just to highlight wrapped output.
