[GRASS-dev] Satellite tiles: tiles missing after import?

Hi,

I imported both the WRS2 Landsat tiles [1] as well as the Sentinel-2
KML [2] (giving extra issues during import + also querying due to
quoting issues).

In both cases, when zooming into the map, some polygons are not drawn
in areas where the tiles massively overlap.

I used the -c flag of v.in.ogr to not break the polygons but to keep
them intact.

Is that a GUI issue or a vector engine issue?

[1] https://landsat.usgs.gov/pathrow-shapefiles --> wrs2_descending.zip
[2] https://sentinel.esa.int/documents/247904/1955685/S2A_OPER_GIP_TILPAR_MPC__20151209T095117_V20150622T000000_21000101T000000_B00.kml

thanks,
Markus

Hi,

2016-12-22 15:25 GMT+01:00 Markus Neteler <neteler@osgeo.org>:

I used the -c flag of v.in.ogr to not break the polygons but to keep
them intact.

number of boundaries and centroids will be correct (29 245). The input
shapefiles contains 28 892 features (some of them are multipolygons,
in the result 29 245 polygons). Building areas from uncleaned
boundaries (basically outer rings) cannot work, Vect_build_line_area
() will not able to report all neighbours. In the results GRASS builds
only 27 751 areas.

Ma

--
Martin Landa
http://geo.fsv.cvut.cz/gwiki/Landa
http://gismentors.cz/mentors/landa

On Thu, Dec 22, 2016 at 6:39 PM, Martin Landa <landa.martin@gmail.com> wrote:

Hi,

2016-12-22 15:25 GMT+01:00 Markus Neteler <neteler@osgeo.org>:

I used the -c flag of v.in.ogr to not break the polygons but to keep
them intact.

number of boundaries and centroids will be correct (29 245). The input
shapefiles contains 28 892 features (some of them are multipolygons,
in the result 29 245 polygons). Building areas from uncleaned
boundaries (basically outer rings) cannot work, Vect_build_line_area()
will not able to report all neighbours. In the results GRASS builds
only 27 751 areas.

So does this affect the display then? I see holes in the map where I
would expect tiles (and QGIS shows them).

thanks
Markus

PS: Sentinel-2 tiles in SHAPE format I found here:
https://github.com/justinelliotmeyers/Sentinel-2-Shapefile-Index
(probably better than the KML)

On Thu, Dec 22, 2016 at 3:25 PM, Markus Neteler <neteler@osgeo.org> wrote:

Hi,

I imported both the WRS2 Landsat tiles [1] as well as the Sentinel-2
KML [2] (giving extra issues during import + also querying due to
quoting issues).

In both cases, when zooming into the map, some polygons are not drawn
in areas where the tiles massively overlap.

I used the -c flag of v.in.ogr to not break the polygons but to keep
them intact.

The -c flag is problematic: it does not break polygons, but polygons are converted to GRASS boundaries. Centroids have to be calculated for each output GRASS area. As a result, an output GRASS area corresponding to an input polygon might have multiple centroids or no centroid at all (these are the holes you see).

Please do not use the -c flag with v.in.ogr unless you know exactly what you are doing and what you will get.

Markus M

Is that a GUI issue or a vector engine issue?

[1] https://landsat.usgs.gov/pathrow-shapefiles → wrs2_descending.zip
[2] https://sentinel.esa.int/documents/247904/1955685/S2A_OPER_GIP_TILPAR_MPC__20151209T095117_V20150622T000000_21000101T000000_B00.kml

thanks,
Markus


grass-dev mailing list
grass-dev@lists.osgeo.org
http://lists.osgeo.org/mailman/listinfo/grass-dev

Hi,

2016-12-22 18:43 GMT+01:00 Markus Neteler <neteler@osgeo.org>:

So does this affect the display then? I see holes in the map where I
would expect tiles (and QGIS shows them).

yes, d.vect in wxGUI displays by default, points, lines, and areas.
See screenshot bellow, in red are boundaries (d.vect type=boundary).
Ma

--
Martin Landa
http://geo.fsv.cvut.cz/gwiki/Landa
http://gismentors.cz/mentors/landa

(attachments)

tiles.png