TileOutsideBounds error with get_geographic_bounds using EPSG:32723 and EPSG:3857 in rio-tiler 7.5.0 #791

CodeWracker · 2025-03-10T13:57:35Z

CodeWracker
Mar 10, 2025

Description
I'm using rio-tiler version 7.5.0 (along with the package versions listed below) with python 3.11.7 and encountering a TileOutsideBounds error when trying to generate a tile from my raster images. I have two orthorectified images with different CRS and bounds (GDrive Link), and both produce a similar error when I use the bounds from src.get_geographic_bounds(crs).

Package Versions

affine = "==2.4.0"
annotated-types = "==0.7.0"
anyio = "==4.8.0"
attrs = "==25.1.0"
blinker = "==1.9.0"
boto3 = "==1.37.7"
botocore = "==1.37.7"
cachetools = "==5.5.2"
certifi = "==2025.1.31"
charset-normalizer = "==3.4.1"
click = "==8.1.8"
click-plugins = "==1.1.1"
cligj = "==0.7.2"
color-operations = "==0.1.6"
flask = "==3.1.0"
h11 = "==0.14.0"
httpcore = "==1.0.7"
httpx = "==0.28.1"
idna = "==3.10"
itsdangerous = "==2.2.0"
jinja2 = "==3.1.6"
jmespath = "==1.0.1"
markupsafe = "==3.0.2"
mercantile = "==1.2.1"
morecantile = "==6.2.0"
numexpr = "==2.10.2"
numpy = "==2.2.3"
opencv-python = "==4.11.0.86"
pillow = "==11.1.0"
pydantic = "==2.10.6"
pydantic-core = "==2.27.2"
pyparsing = "==3.2.1"
pyproj = "==3.7.1"
pystac = "==1.12.2"
python-dateutil = "==2.9.0.post0"
rasterio = "==1.4.3"
requests = "==2.32.3"
rio-color = "==2.0.1"
rio-mucho = "==1.0.0"
rio-tiler = "==7.5.0"
s3transfer = "==0.11.4"
six = "==1.17.0"
sniffio = "==1.3.1"
typing-extensions = "==4.12.2"
urllib3 = "==2.3.0"
werkzeug = "==3.1.3"
matplotlib = "*"

Steps to Reproduce

I have two orthorectified images:
- Image 1 (EPSG:32723):
  - Bounds: (310673.97175115923, 7655637.655109174, 310952.8864424243, 7656025.083149253)
- Image 2 (EPSG:3857):
  - Bounds: (-5212400.40139766, -2414656.836298912, -5212096.667121954, -2414235.643755028)

I used the following code (taken from a GeoPythonConference video at 8:28):

import io
import numpy as np
from PIL import Image
from rio_tiler.io import Reader
from rio_tiler.io import rasterio
from pyproj import Transformer
from matplotlib.pyplot import imshow

with Reader('orto-mercator.tif') as src:
    crs = src.crs
    print("crs", crs)
    bounds = src.get_geographic_bounds(crs)
    print("bounds", bounds)
    print("minzoom", src.minzoom)
    tile = src.tms.tile
    print(tile)
    img = src.tile(bounds[0], bounds[3], src.minzoom + 1)
    imshow(img)

Output / Error for EPSG:3857 (Mercator):

crs EPSG:3857
bounds (-5212400.40139766, -2414656.836298912, -5212096.667121954, -2414235.643755028)
minzoom 16
<bound method TileMatrixSet.tile of <TileMatrixSet title='Google Maps Compatible for the World' id='WebMercatorQuad' crs='http://www.opengis.net/def/crs/EPSG/0/3857'>>
Traceback (most recent call last):
  ...
rio_tiler.errors.TileOutsideBounds: Tile(x=-5212400.40139766, y=-2414235.643755028, z=17) is outside bounds

Output / Error for EPSG:32723 (Lat/Long):

crs EPSG:32723
bounds (310673.97175115923, 7655637.655109174, 310952.8864424243, 7656025.083149253)
minzoom 16
<bound method TileMatrixSet.tile of <TileMatrixSet title='Google Maps Compatible for the World' id='WebMercatorQuad' crs='http://www.opengis.net/def/crs/EPSG/0/3857'>>
Traceback (most recent call last):
  ...
rio_tiler.errors.TileOutsideBounds: Tile(x=310673.97175115923, y=7656025.083149253, z=17) is outside bounds

In both cases, regardless of whether I use the Mercator or the lat/long image, I always get the error stating that the tile is outside of bounds—even though the coordinates are derived directly from the source.

Expected Behavior
I expected to generate a valid tile based on the bounds returned by src.get_geographic_bounds(crs). The error suggests that the calculated tile coordinates are outside the valid range, but these coordinates are directly taken from the source data.

Questions / Request for Guidance

Could someone explain why the tile generated from these bounds is considered out of bounds?
Am I misinterpreting how to generate the tile or the meaning of the bounds provided?
Is there a recommended approach to correctly generate a tile using these bounds?

Any guidance or suggestions on how to resolve this issue would be greatly appreciated.

Thanks in advance for your help!

Answered by CodeWracker

Mar 10, 2025

There is a problem with this solution, but you put me in the right track!

your suggestion throws

C:\Users\ralph\.virtualenvs\teste-nEKyzodD\Lib\site-packages\rio_tiler\io\rasterio.py:135: NoOverviewWarning: The dataset has no Overviews. rio-tiler performances might be impacted.
  warnings.warn(
bounds (-46.823789475262465, -21.190907136062886, -46.82106098384069, -21.18737930203305)
minzoom 16
Tile(x=48487, y=73432, z=17)
Traceback (most recent call last):
  File "E:\projetos\teste\main.py", line 44, in <module>
    imshow(img)
  File "C:\Users\ralph\.virtualenvs\teste-nEKyzodD\Lib\site-packages\matplotlib\pyplot.py", line 3590, in imshow
    __ret = gca().imshow(
            ^^^^^^^^^^^^^
…

View full answer

vincentsarago · 2025-03-10T14:03:20Z

vincentsarago
Mar 10, 2025
Maintainer

@CodeWracker

with Reader('orto-mercator.tif') as src:
    crs = src.crs
    print("crs", crs)
    bounds = src.get_geographic_bounds(crs)

here you're using the wrong CRS. You want to use the geographic_crs used by the TileMatrixSet (WebMercatorQuad by default)

import io
import numpy as np
from PIL import Image
from rio_tiler.io import Reader
from rio_tiler.io import rasterio
from pyproj import Transformer
from matplotlib.pyplot import imshow

with Reader('orto-mercator.tif') as src:
    bounds = src.get_geographic_bounds(src.tms.rasterio_geographic_crs)
    print("bounds", bounds)
    print("minzoom", src.minzoom)
    tile = src.tms.tile
    print(tile)
    img = src.tile(bounds[0], bounds[3], src.minzoom + 1)
    imshow(img)

5 replies

CodeWracker Mar 10, 2025
Author

Nops, same error with you code suggestion

PS E:\projetos\teste> python .\main.py
C:\Users\ralph\.virtualenvs\teste-nEKyzodD\Lib\site-packages\rio_tiler\io\rasterio.py:135: NoOverviewWarning: The dataset has no Overviews. rio-tiler performances might be impacted.
  warnings.warn(
bounds (-46.823789475262465, -21.190907136062886, -46.82106098384069, -21.18737930203305)
minzoom 16
<bound method TileMatrixSet.tile of <TileMatrixSet title='Google Maps Compatible for the World' id='WebMercatorQuad' crs='http://www.opengis.net/def/crs/EPSG/0/3857>>
Traceback (most recent call last):
  File "E:\projetos\teste\main.py", line 15, in <module>
    img = src.tile(bounds[0], bounds[3], src.minzoom + 1)
          ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "C:\Users\ralph\.virtualenvs\teste-nEKyzodD\Lib\site-packages\rio_tiler\io\rasterio.py", line 279, in tile
    raise TileOutsideBounds(
rio_tiler.errors.TileOutsideBounds: Tile(x=-46.823789475262465, y=-21.18737930203305, z=17) is outside bounds

PS E:\projetos\teste> python .\main.py
C:\Users\ralph\.virtualenvs\teste-nEKyzodD\Lib\site-packages\rio_tiler\io\rasterio.py:135: NoOverviewWarning: The dataset has no Overviews. rio-tiler performances might be impacted.
  warnings.warn(
bounds (-46.82378947526247, -21.190907076244944, -46.82106082776861, -21.187379302033055)
minzoom 16
<bound method TileMatrixSet.tile of <TileMatrixSet title='Google Maps Compatible for the World' id='WebMercatorQuad' crs='http://www.opengis.net/def/crs/EPSG/0/3857>>
Traceback (most recent call last):
  File "E:\projetos\teste\main.py", line 15, in <module>
    img = src.tile(bounds[0], bounds[3], src.minzoom + 1)
          ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "C:\Users\ralph\.virtualenvs\teste-nEKyzodD\Lib\site-packages\rio_tiler\io\rasterio.py", line 279, in tile
    raise TileOutsideBounds(
rio_tiler.errors.TileOutsideBounds: Tile(x=-46.82378947526247, y=-21.187379302033055, z=17) is outside bounds

(same error for both files)

vincentsarago Mar 10, 2025
Maintainer

😓 sorry I copy/pasted the same error

you need to pass tile indexes not the bounds coordinates

import io
import numpy as np
from PIL import Image
from rio_tiler.io import Reader
from rio_tiler.io import rasterio
from pyproj import Transformer
from matplotlib.pyplot import imshow

with Reader('orto-mercator.tif') as src:
    bounds = src.get_geographic_bounds(src.tms.rasterio_geographic_crs)
    print("bounds", bounds)
    print("minzoom", src.minzoom)
    tile = src.tms.tile(bounds[0], bounds[3], src.minzoom + 1)
    print(tile)
    img = src.tile(*tile)
    imshow(img)

CodeWracker Mar 10, 2025
Author

There is a problem with this solution, but you put me in the right track!

your suggestion throws

C:\Users\ralph\.virtualenvs\teste-nEKyzodD\Lib\site-packages\rio_tiler\io\rasterio.py:135: NoOverviewWarning: The dataset has no Overviews. rio-tiler performances might be impacted.
  warnings.warn(
bounds (-46.823789475262465, -21.190907136062886, -46.82106098384069, -21.18737930203305)
minzoom 16
Tile(x=48487, y=73432, z=17)
Traceback (most recent call last):
  File "E:\projetos\teste\main.py", line 44, in <module>
    imshow(img)
  File "C:\Users\ralph\.virtualenvs\teste-nEKyzodD\Lib\site-packages\matplotlib\pyplot.py", line 3590, in imshow
    __ret = gca().imshow(
            ^^^^^^^^^^^^^
  File "C:\Users\ralph\.virtualenvs\teste-nEKyzodD\Lib\site-packages\matplotlib\__init__.py", line 1521, in inner
    return func(
           ^^^^^
  File "C:\Users\ralph\.virtualenvs\teste-nEKyzodD\Lib\site-packages\matplotlib\axes\_axes.py", line 5976, in imshow
    im.set_data(X)
  File "C:\Users\ralph\.virtualenvs\teste-nEKyzodD\Lib\site-packages\matplotlib\image.py", line 685, in set_data
    self._A = self._normalize_image_array(A)
              ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "C:\Users\ralph\.virtualenvs\teste-nEKyzodD\Lib\site-packages\matplotlib\image.py", line 648, in _normalize_image_array
    raise TypeError(f"Image data of dtype {A.dtype} cannot be "
TypeError: Image data of dtype object cannot be converted to float

trying to fix it with

import io
import numpy as np
from PIL import Image
from rio_tiler.io import Reader
from rio_tiler.io import rasterio
from pyproj import Transformer
from matplotlib.pyplot import imshow, show

with Reader('orto-mercator.tif') as src:
    bounds = src.get_geographic_bounds(src.tms.rasterio_geographic_crs)
    print("bounds", bounds)
    print("minzoom", src.minzoom)
    center_lon = (bounds[0] + bounds[2]) / 2
    center_lat = (bounds[1] + bounds[3]) / 2
    zoom = 22
    tile = src.tms.tile(center_lon, center_lat, src.minzoom + 1)
    print(tile)
    img, mask = src.tile(*tile)
    # Transpõe o array se ele estiver no formato (bandas, altura, largura)
    img = np.moveaxis(img, 0, -1)
    
    imshow(img)
    show()

and it works!! (changed the position lon, lat to the center to better visualize the image)

for reference, if it is needed to use mercantile crs the solution would be

import mercantile
import numpy as np
from rio_tiler.io import Reader
from matplotlib.pyplot import imshow, show

with Reader("orto.tif") as src:
    # Obtém os limites na CRS que o TMS espera (normalmente EPSG:4326)
    bounds = src.get_geographic_bounds(src.tms.rasterio_geographic_crs)
    # Calcula o ponto central dos limites
    center_lon = (bounds[0] + bounds[2]) / 2
    center_lat = (bounds[1] + bounds[3]) / 2
    zoom = 22
    print("Ponto central:", center_lon, center_lat)
    print("Zoom:", zoom)

    # Converte o ponto central em índices de tile
    tile_idx = mercantile.tile(center_lon, center_lat, zoom)
    print("Índices do tile:", tile_idx)
    
    # Desempacota o resultado do método tile() (data, mask)
    img, mask = src.tile(tile_idx.x, tile_idx.y, tile_idx.z)
    
    # Transpõe o array se ele estiver no formato (bandas, altura, largura)
    img = np.moveaxis(img, 0, -1)
    
    imshow(img)
    show()

Answer selected by vincentsarago

CodeWracker Mar 10, 2025
Author

THANK YOU! <3

vincentsarago Mar 10, 2025
Maintainer

FYI, morecantile is just like mercantile, except it can work with more TMS

tms = morecantile.tms.get("WebMercatorQuad") is like import mercantile as tms

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

TileOutsideBounds error with get_geographic_bounds using EPSG:32723 and EPSG:3857 in rio-tiler 7.5.0 #791

{{title}}

{{editor}}'s edit

{{editor}}'s edit

Replies: 1 comment 5 replies

{{title}}

{{title}}

{{editor}}'s edit

{{editor}}'s edit

{{title}}

{{title}}

{{title}}

{{title}}

Select a reply

TileOutsideBounds error with get_geographic_bounds using EPSG:32723 and EPSG:3857 in rio-tiler 7.5.0 #791

CodeWracker Mar 10, 2025

Replies: 1 comment · 5 replies

vincentsarago Mar 10, 2025 Maintainer

CodeWracker Mar 10, 2025 Author

vincentsarago Mar 10, 2025 Maintainer

CodeWracker Mar 10, 2025 Author

CodeWracker Mar 10, 2025 Author

vincentsarago Mar 10, 2025 Maintainer

CodeWracker
Mar 10, 2025

Replies: 1 comment 5 replies

vincentsarago
Mar 10, 2025
Maintainer

CodeWracker Mar 10, 2025
Author

vincentsarago Mar 10, 2025
Maintainer

CodeWracker Mar 10, 2025
Author

CodeWracker Mar 10, 2025
Author

vincentsarago Mar 10, 2025
Maintainer