Lupang Arenda, Santa Ana, Taytay community mapping collaboration

[maning]

We had a great fieldmapping at Lupang Arenda today! Thanks to all who volunteered!
See this ticket for context of the collaboration.

The crew

[WIP] left-Bing imagery, right-sample orthophoto from drone captures

I'm outlining some key actions to people involved, let's use this ticket as public record of the further work.

• Upload street-level photos to Mapillary @govvin
• Test processing of drone photos with OpenDroneMap @maning
• Compare quality of orthophoto from drones with ODM process and other proprietary tools Bunny and Leigh
• Upload orthophoto to OpenAerialMap @banitopinoy
• Create PR to imagery-index to load Arenda Orthophoto from OAM as default imagery for OSM editors @maning
• Share all output to community contacts @govvin
• Scope next actions with community after delivery of drone maps @govvin @ralleon

[maning]

Processing drone imagery with OpenDroneMap

• Temporary download link until final processed is uploaded to OAM (814 MB): https://www.dropbox.com/sh/m7msxaup2gtbryj/AACOqtvrmrxHwOGE7xyjOk3Va?dl=0
• Using the docker container of OpenDroneMap, I processed all 5 flights. Average processing time is 3.3 hours each.
• Processing is straightforward if you have docker installed in your computer. All you need is a directory of the imagery and run the docker container. One important change in your docker preferences is to increase RAM to at least 8 GB.
• Orthophoto output is in UTM Z51N.

gdalinfo /Users/maningsambale/Projects/odm/bunny1_orthophoto/odm_orthophoto.original.tif
Driver: GTiff/GeoTIFF
Files: /Users/maningsambale/Projects/odm/bunny1_orthophoto/odm_orthophoto.original.tif
Size is 16589, 9006
Coordinate System is:
PROJCS["WGS 84 / UTM zone 51N",
    GEOGCS["WGS 84",
        DATUM["WGS_1984",
            SPHEROID["WGS 84",6378137,298.257223563,
                AUTHORITY["EPSG","7030"]],
            AUTHORITY["EPSG","6326"]],
        PRIMEM["Greenwich",0,
            AUTHORITY["EPSG","8901"]],
        UNIT["degree",0.0174532925199433,
            AUTHORITY["EPSG","9122"]],
        AUTHORITY["EPSG","4326"]],
    PROJECTION["Transverse_Mercator"],
    PARAMETER["latitude_of_origin",0],
    PARAMETER["central_meridian",123],
    PARAMETER["scale_factor",0.9996],
    PARAMETER["false_easting",500000],
    PARAMETER["false_northing",0],
    UNIT["metre",1,
        AUTHORITY["EPSG","9001"]],
    AXIS["Easting",EAST],
    AXIS["Northing",NORTH],
    AUTHORITY["EPSG","32651"]]
Origin = (296108.531677999999374,1607653.094869999913499)
Pixel Size = (0.049998330399662,-0.049995508549838)
Metadata:
  AREA_OR_POINT=Area
Image Structure Metadata:
  COMPRESSION=DEFLATE
  INTERLEAVE=PIXEL
Corner Coordinates:
Upper Left  (  296108.532, 1607653.095) (121d 6'28.09"E, 14d32' 3.14"N)
Lower Left  (  296108.532, 1607202.835) (121d 6'28.22"E, 14d31'48.49"N)
Upper Right (  296937.954, 1607653.095) (121d 6'55.79"E, 14d32' 3.36"N)
Lower Right (  296937.954, 1607202.835) (121d 6'55.92"E, 14d31'48.71"N)
Center      (  296523.243, 1607427.965) (121d 6'42.01"E, 14d31'55.92"N)
Band 1 Block=512x512 Type=Byte, ColorInterp=Red
  Mask Flags: PER_DATASET ALPHA
Band 2 Block=512x512 Type=Byte, ColorInterp=Green
  Mask Flags: PER_DATASET ALPHA
Band 3 Block=512x512 Type=Byte, ColorInterp=Blue
  Mask Flags: PER_DATASET ALPHA
Band 4 Block=512x512 Type=Byte, ColorInterp=Alpha

• Over to Bunny and Leigh to compare quality of processing with other tools.

Details of processing time below:

command	# images	size (MB)	runtime
time docker run -it --rm -v "$(pwd)/bunny1:/code/images" -v "$(pwd)/bunny1_orthophoto:/code/odm_orthophoto" opendronemap/odm	70	350	real 102m58.414s user 0m0.280s sys 0m0.215s
time docker run -it --rm -v "$(pwd)/bunny3:/code/images" -v "$(pwd)/bunny3_orthophoto:/code/odm_orthophoto" opendronemap/odm	41	205	real 194m11.746s user 0m0.314s sys 0m0.361s
time docker run -it --rm -v "$(pwd)/bunny2:/code/images" -v "$(pwd)/bunny2_orthophoto:/code/odm_orthophoto" opendronemap/odm	55	270	real 256m1.920s user 0m0.436s sys 0m0.571s
time docker run -it --rm -v "$(pwd)/bunny4:/code/images" -v "$(pwd)/bunny4_orthophoto:/code/odm_orthophoto" opendronemap/odm	68	340	real 219m44.133s user 0m0.501s sys 0m0.668s
time docker run -it --rm -v "$(pwd)/bunny5:/code/images" -v "$(pwd)/bunny4_orthophoto:/code/odm_orthophoto" opendronemap/odm	90	450	real 224m35.409s user 0m0.523s sys 0m0.685s

[govvin]

Thank you for the update @maning and the tips for tweaking the Docker settings.

We collected 2,600 images but only 1,600 images uploaded to Mapillary, after removing duplicates.

[govvin]

Diary post by Maning: https://www.openstreetmap.org/user/maning/diary/390221

[maning]

Imagery now available in OAM: https://map.openaerialmap.org/#/121.11478328704835,14.537581224033948,15?resolution=high&_k=dqxutz

[banitopinoy]

The available flyover images in OpenAerialMap can now be used as WMS in QGIS

OAM imagery can also used directly as background in OSM ID editor or JOSM

[govvin]

@banitopinoy , could you publish the stitched images to OAM? We could start doing mapathons to digitize the features in the area with a Tasking Manager (to avoid editing conflicts) while waiting for the completion of the remaining areas.

[banitopinoy]

@banitopinoy , could you publish the stitched images to OAM? We could start doing mapathons to digitize the features in the area with a Tasking Manager (to avoid editing conflicts) while waiting for the completion of the remaining areas.

All images are already available in OpenAerialMap:

[govvin]

Our CWTS volunteers from PUPSJ will be contributing to the Lupang Arenda mapping effort using Task 6228. Feel free to share the task, and lend a hand.

[govvin]

@banitopinoy I meant a stitched collection of all those imagery.

[banitopinoy]

The first image in the left corner is the stiched image.

[govvin]

@banitopinoy Gaaahh. I need new eyeglasses. Sorry for missing that.

[maning]

Draft imagery-index here: https://gist.github.com/maning/93fe6cd3ad3529eee7a95354ec99e63e
These are the allowed values in the schema: https://github.com/osmlab/editor-layer-index/blob/gh-pages/schema.json

[maning]

Create PR to imagery-index to load Arenda Orthophoto from OAM as default imagery for OSM editors @maning

Done! https://osmlab.github.io/editor-layer-index/

[maning]

Reposting here @ralleon assessment on image alignment.

With Strava cycle heatmap on top of the "draft" drone imagery:
Best-fitting heatmap on minor roads, with consideration of cyclist-bias (of not using the cyclelanes) on C6 Road, the general imagery offset was estimated to be around 0.74, -2.76 (with approx 0.8 meter repeatable error)

[maning]

It's now available in iD master! Once a new release of iD becomes available to osm.org, this will be the top recommended imagery.

[maning]

This is now available in osm.org's iD (v.2.15.5).

Are there any next actions in this ticket or good to close?

[govvin]

@maning Finding a workable schedule for the Arenda community and the drone pilot to cover remaining area of Arenda.

How's your schedule these days @leighmedy ?

[maning]

We did round 2 mission last Oct 12. Data is now available in OAM and soon directly in iD

This mission completes the coverage of the whole AOI. Thanks to all involved!

Photos!

[maning]

PR to update the coverage of Arenda in the editor-layer-index was rejected. I will start a fresh PR after getting clarity on the process to be outlined by the repo maintainers.

[maning]

Marianne and Kath would like to print a map of CELO coverage within Lupang Arenda. They will use the map to showcase the progress of mapping to the community. The CELO mapped buildings are here.

Requirements: Poster size or larger (2x3 or 4x8 feet) map showing the CELO tagged buildings, drone imagery, additional context like roads, purok boundaries.

telegram chat:

[banitopinoy]

Possible to print in large format and it might be more practical if it will printed on a tarpaulin. A 4x8ft print would cost around P500. There is a good print shop in Concepcion Uno in Marikina which I have tried that can do the job. Main constraint is that we have this current pandemic concern. MM is lock down and print shops are temporarily close.

[govvin]

Pending tasks

• Building tracing (as of 3/20, 100% complete. )
• Validate tasks (as of 3/23, only 6% validated.) <-- DO NOT SKIP THIS
• CELO validation (as of 3/20, 93% complete
• review and validate boundary tags and elements used by Marianne.
• prepare layouts for printable maps (by purok? all of L.A.?)