Too many points (10201 out of 10201) failed to transform, unable to compute output bounds

Question

I am trying to read the drone imagery (20 cm) with rasterio library but getting an error.

Traceback (most recent call last):
  File "/home/manap/PycharmProjects/rio-tiler/multi.py", line 33, in <module>
    with COGReader(r"/home/manap/munich_1.tif") as image:
  File "<attrs generated init rio_tiler.io.cogeo.COGReader>", line 14, in __init__
  File "/home/manap/PycharmProjects/rio-tiler/rio_tiler/io/cogeo.py", line 107, in __attrs_post_init__
    self._set_zooms()
  File "/home/manap/PycharmProjects/rio-tiler/rio_tiler/io/cogeo.py", line 155, in _set_zooms
    minzoom, maxzoom = self.get_zooms()
  File "/home/manap/PycharmProjects/rio-tiler/rio_tiler/io/cogeo.py", line 137, in get_zooms
    *self.dataset.bounds,
  File "/home/manap/PycharmProjects/rio-tiler/venv/lib/python3.6/site-packages/rasterio/env.py", line 387, in wrapper
    return f(*args, **kwds)
  File "/home/manap/PycharmProjects/rio-tiler/venv/lib/python3.6/site-packages/rasterio/warp.py", line 509, in calculate_default_transform
    src_crs, dst_crs, width, height, left, bottom, right, top, gcps, rpcs, **kwargs
  File "rasterio/_warp.pyx", line 697, in rasterio._warp._calculate_default_transform
  File "rasterio/_warp.pyx", line 679, in rasterio._warp._calculate_default_transform
  File "rasterio/_err.pyx", line 192, in rasterio._err.exc_wrap_int
rasterio._err.CPLE_AppDefinedError: Too many points (10201 out of 10201) failed to transform, unable to compute output bounds.
Process finished with exit code 1

my code:

from rio_tiler.io import COGReader
import rio_tiler
with COGReader(r"/home/manap/munich_1.tif") as image:
    print(image)

the image can be downloaded from https://eoxplore-my.sharepoint.com/:f:/g/personal/chiefscienceofficer_eoxplore_onmicrosoft_com/EpsWYilRNh5Eld4cU5p_fe4BVQX7FBoRMZXnD89qKba80w

huckfinn · Answer 1 · 2021-12-10T10:30:23.440

I cannot find an munich_1.tif. Instead I found a image DOP_20/4468260_5331209.tif in your repository. The image has a dimension of 16425x12760:

gdalinfo 4468260_5331209.tif 
Driver: GTiff/GeoTIFF
Files: 4468260_5331209.tif
       4468260_5331209.tif.ovr
Size is 16425, 12760
Coordinate System is:
PROJCS["DHDN / 3-degree Gauss-Kruger zone 4",
    GEOGCS["DHDN",
        DATUM["Deutsches_Hauptdreiecksnetz",
            SPHEROID["Bessel 1841",6377397.155,299.1528128,
                AUTHORITY["EPSG","7004"]],
            TOWGS84[598.1,73.7,418.2,0.202,0.045,-2.455,6.7],
            AUTHORITY["EPSG","6314"]],
        PRIMEM["Greenwich",0,
            AUTHORITY["EPSG","8901"]],
        UNIT["degree",0.0174532925199433,
            AUTHORITY["EPSG","9122"]],
        AUTHORITY["EPSG","4314"]],
    PROJECTION["Transverse_Mercator"],
    PARAMETER["latitude_of_origin",0],
    PARAMETER["central_meridian",12],
    PARAMETER["scale_factor",1],
    PARAMETER["false_easting",4500000],
    PARAMETER["false_northing",0],
    UNIT["metre",1,
        AUTHORITY["EPSG","9001"]],
    AUTHORITY["EPSG","31468"]]
Origin = (4468259.984000000171363,5333761.273000000044703)
Pixel Size = (0.200000000000000,-0.200000000000000)
Metadata:
  AREA_OR_POINT=Area
  TIFFTAG_COPYRIGHT=*************************
  TIFFTAG_DATETIME=07.05.2018 00:00:00
  TIFFTAG_RESOLUTIONUNIT=3 (pixels/cm)
  TIFFTAG_SOFTWARE=*************************
  TIFFTAG_XRESOLUTION=250
  TIFFTAG_YRESOLUTION=250
Image Structure Metadata:
  INTERLEAVE=PIXEL
Corner Coordinates:
Upper Left  ( 4468259.984, 5333761.273) ( 11d34'24.42"E, 48d 8'31.22"N)
Lower Left  ( 4468259.984, 5331209.273) ( 11d34'25.11"E, 48d 7' 8.59"N)
Upper Right ( 4471544.984, 5333761.273) ( 11d37' 3.35"E, 48d 8'31.78"N)
Lower Right ( 4471544.984, 5331209.273) ( 11d37' 3.96"E, 48d 7' 9.15"N)
Center      ( 4469902.484, 5332485.273) ( 11d35'44.21"E, 48d 7'50.19"N)
Band 1 Block=16425x8 Type=Byte, ColorInterp=Red
  Overviews: 8213x6380, 4107x3190, 2054x1595, 1027x798, 514x399, 257x200
Band 2 Block=16425x8 Type=Byte, ColorInterp=Green
  Overviews: 8213x6380, 4107x3190, 2054x1595, 1027x798, 514x399, 257x200
Band 3 Block=16425x8 Type=Byte, ColorInterp=Blue
  Overviews: 8213x6380, 4107x3190, 2054x1595, 1027x798, 514x399, 257x200

According to the code in warp.py the operation is limited to a image size of 100000x100000 cells.

"""$ rio warp"""
...
Improper usage of rio-warp can lead to accidental creation of
extremely large datasets. We'll put a hard limit on the size of
datasets and raise a usage error if the limits are exceeded.
MAX_OUTPUT_WIDTH = 100000
MAX_OUTPUT_HEIGHT = 100000

So the source dimension 16425x12760 should not be an issue. When I install the environment on Ubuntu 18.04 - 64 bit with

  pip3 install rasterio 
  pip3 install rio_tiler

,download the DOP_20 image and call my the example code

#!/usr/bin/env python3
from rio_tiler.io import COGReader
import rio_tiler
FILE='DOP_20/4468260_5331209.tif'
with COGReader(FILE) as image:
    print(image)

I get no error during the image open step.

COGReader(\
   bounds=(11.572063386087754, 48.11813315397459, \
           11.61637335456606, 48.14123922531762), \
   filepath='DOP_20/4468260_5331209.tif', \
   dataset=<open DatasetReader \ 
   name='DOP_20/4468260_5331209.tif' \
   mode='r'>, \
   tms=<TileMatrixSet \
        title='Google Maps Compatible for the World' \
        identifier='WebMercatorQuad'>, \
   minzoom=13, \
   maxzoom=19, \
   colormap={}, \
   nodata=None, \
   unscale=None, \
   resampling_method=None, \
   vrt_options=None, \
   post_process=None, \
   _kwargs={})

IMO you try to transform the image, with some code not explained in your question. Can you provide this code and the parameter set for the warp call.

Review

I assume you have copied the image DOP_20/4468260_5331209.tif else where and renamed it. The image has some more metadata in the directory DOP_20:

4468260_5331209.tfw
4468260_5331209.tif
4468260_5331209.tif.ovr
4468260_5331209.txt
4468260_5331209.wld

I've read the data from the original DOP_20 directory and got no error, while your copy make some trouble with the rastreio. To make the Geotif portable and self contained with the overview images, you could use the commands gdal_translate and gdaladdo:

gdal_translate -co BIGTIF=YES      \
               -co COMPRESS=LZW    \
               -co SPARSE_OK=TRUE  \
               -co TILED=yes       \
               -co BLOCKXSIZE=1024 \
               -co BLOCKYSIZE=1024 \
               4468260_5331209.tif \ 
               4468260_5331209.blocks.tif
gdaladdo -r lanczos 4468260_5331209.blocks.tif 2 4 8 16 32 64

When I read your script right, then you try to split the image into 256x256 tiles driven by your text file. The script triggers for all entries exceptions the exception TileOutsideBounds.

~/Source/IfGDV-GSE/Test-Tile/tiling$ ./script.py
278707
181023
EXCEPT: TileOutsideBounds
...

Here is your adopted script:

#!/usr/bin/env python3
from rio_tiler.io import COGReader
import os
import rio_tiler
import pandas as pd
import time
from PIL import Image
Preparation -------------------------------------
PRJ_DIR=$HOME/Source/IfGDV-GSE/Test-Tile
mv tiling-20211210T084355Z-001.zip $PRJ_DIR
unzip $PRJ_DIR/tiling-20211210T084355Z-001.zip
cd $PRJ_DIR/tiling
mkdir result
gdal_translate -co BIGTIF=YES -co COMPRESS=LZW -co SPARSE_OK=TRUE -co TILED=yes -co BLOCKXSIZE=1024 -co BLOCKYSIZE=1024 munich_20cm.tif munich_20cm.blocks.tif
gdaladdo -r lanczos munich_20cm.blocks.tif 2 4 8 16 32 64
mv script.txt script.py
dos2unix  script.py
chmod 755 script.py
emacs -nw script.py
--------------------------------------------------
add shebang #!/usr/bin/env python3
add some variables
HOME        = os.environ.get('HOME')
WORK_DIR    = HOME+'/Source/IfGDV-GSE/Test-Tile/tiling/'
DST_PATH    = WORK_DIR+'result/'
SRC_LIST    = WORK_DIR+'new_munich.txt'
SRC_IMAGE = WORK_DIR+'munich_20cm.tif'
SRC_IMAGE   = WORK_DIR+'munich_20cm.blocks.tif'
SRC_FILE   = open(SRC_LIST, 'r')
SRC_POINTS = SRC_FILE.readlines()
image =  COGReader(SRC_IMAGE)
number = len(SRC_POINTS)
for coordinates in SRC_POINTS:
x = int(coordinates.split(',')[0])
y = int(coordinates.split(',')[1])
print(x)
print(y)
z = 19
number -= 1
try:
    filename = DST_PATH + str(x) +'/' + str(y) + '.png'
    print(filename)
    img = image.tile(x, y, z, tilesize=256)
    buff = img.render()

    # check for existence of the direction
    x_direction = os.path.join(DST_PATH, str(x))
    if os.path.exists(x_direction) == False:
        os.mkdir(x_direction)
    with open(os.path.join(x_direction, str(y) + '.png'), &quot;wb&quot;) as f:
        f.write(buff)

    #os.remove(filename)

except rio_tiler.errors.TileOutsideBounds:
    print(&quot;EXCEPT: TileOutsideBounds&quot;)

thanks for your answer. I am trying only to read the image but I am getting an error. Have you done some edits with the image or code? — Manap Shymyr, Dec 09 '21 at 08:00
I've installed rasterio and and rio_tiler via pip3 on Ubuntu 18.04, and changed as mentioned above the file name because munich_1.tif was not in the DOP_20 directory. Where does your munich_1.tif comes from? Is it the same files a DOP_20/4468260_5331209.tif or is it a larger file not in provided DOP_20 remote directory? — huckfinn, Dec 09 '21 at 08:37
And the size is the same? Can you check with gdalinfo munich_1.tif. — huckfinn, Dec 09 '21 at 11:21
If I send x,y,z coordinates as a txt file can you please tile it for me? the script is read, you just need to configure directories. — Manap Shymyr, Dec 10 '21 at 07:41
i have uploaded files to here: https://drive.google.com/drive/folders/1zneVCdlIIAEjyQvnKZjCHCs9MMQMxUTu?usp=sharing — Manap Shymyr, Dec 10 '21 at 08:12
what if you try to use an initial image which is DOP_20/4468260_5331209.tif — Manap Shymyr, Dec 10 '21 at 09:26

Too many points (10201 out of 10201) failed to transform, unable to compute output bounds

1 Answers1

Improper usage of rio-warp can lead to accidental creation of

extremely large datasets. We'll put a hard limit on the size of

datasets and raise a usage error if the limits are exceeded.

Review

Preparation -------------------------------------

PRJ_DIR=$HOME/Source/IfGDV-GSE/Test-Tile

mv tiling-20211210T084355Z-001.zip $PRJ_DIR

unzip $PRJ_DIR/tiling-20211210T084355Z-001.zip

cd $PRJ_DIR/tiling

mkdir result

gdal_translate -co BIGTIF=YES -co COMPRESS=LZW -co SPARSE_OK=TRUE -co TILED=yes -co BLOCKXSIZE=1024 -co BLOCKYSIZE=1024 munich_20cm.tif munich_20cm.blocks.tif

gdaladdo -r lanczos munich_20cm.blocks.tif 2 4 8 16 32 64

mv script.txt script.py

dos2unix script.py

chmod 755 script.py

emacs -nw script.py

--------------------------------------------------

add shebang #!/usr/bin/env python3

add some variables

SRC_IMAGE = WORK_DIR+'munich_20cm.tif'

Linked