Image.open 报错 Cannot identify image file
Image.open gives error Cannot identify image file
我正在尝试使用 PIL 的图像功能打开一个 geotiff 文件。它引发错误：
OSError: cannot identify image file 'Whatever\image\I\use.tiff'
我看到问题被问到 here 例如，解决方案要么使用
Import Image
而不是
From PIL import Image
我认为这是一个过时的解决方案；我无法导入图像。另一个解决方案是将 pillow 更新到 2.9，但 5 年后我们使用的是 5.0.0。我也尝试了 4.0.0，但我收到了同样的错误。
是否有最新的解决方案？
这是我的代码，here 是一个 link 文件：
image_path = 'each\image\I\use.tiff'
from PIL import Image
Image.open(image_path)
您有一个 multi-channel 32 位 TIFF 图像，Pillow 尚不支持该格式。见 issue #1888:
Pillow (and PIL) is currently able to open 8 bit per channel multi-channel images (such as RGB) but is able to open higher bit depth images (e.g. I16, I32, or Float32 images) if they are single channel (e.g., grayscale).
[...]
Requirements
- We should be able to support common GIS formats as well as high bit depth RGB(A) images.
- At least 4 channels, but potentially more (see #1839)
- Different pixel formats, including I16, I32, and Float.
我通过直接使用 TIFF 插件图像 reader 确定了这一点，并启用了调试模式：
from PIL import TiffImagePlugin
TiffImagePlugin.DEBUG = True
with open(image_path, 'rb') as f:
TiffImagePlugin.TiffImageFile(f)
其中包括输出：
tag: BitsPerSample (258) - type: short (3) Tag Location: 46 - Data Location: 218 - value: (32, 32, 32)
（下面是完整的调试输出）
您可以使用 Python GDAL bindings to read this format. You can also use the gdal_translate command line utility 将您的文件转换为 Pillow 可以 处理的格式；对于多波段，您必须降低到 8 位，或者移动到灰度。
例如，要将您的输入文件转换为 PNG，您可以使用：
gdal_translate -of PNG rgb_CGI.tiff rgb_CGI.png
之后 Pillow 可以打开 PNG 文件。
Pillow 的 TIFF 插件的完整调试输出：
>>> from PIL import TiffImagePlugin
>>> TiffImagePlugin.DEBUG = True
>>> with open(image_path, 'rb') as f:
... TiffImagePlugin.TiffImageFile(f)
...
*** TiffImageFile._open ***
- __first: 8
- ifh: b'II*\x00\x08\x00\x00\x00'
Seeking to frame 0, on frame -1, __next 8, location: 8
Loading tags, location: 8
tag: ImageWidth (256) - type: short (3) - value: 2924
tag: ImageLength (257) - type: short (3) - value: 2088
tag: BitsPerSample (258) - type: short (3) Tag Location: 46 - Data Location: 218 - value: (32, 32, 32)
tag: Compression (259) - type: short (3) - value: 1
tag: PhotometricInterpretation (262) - type: short (3) - value: 1
tag: StripOffsets (273) - type: long (4) Tag Location: 82 - Data Location: 8576 - value: <table: 8352 bytes>
tag: SamplesPerPixel (277) - type: short (3) - value: 3
tag: RowsPerStrip (278) - type: short (3) - value: 1
tag: StripByteCounts (279) - type: long (4) Tag Location: 118 - Data Location: 224 - value: <table: 8352 bytes>
tag: PlanarConfiguration (284) - type: short (3) - value: 1
tag: ExtraSamples (338) - type: short (3) - value: (0, 0)
tag: SampleFormat (339) - type: short (3) Tag Location: 154 - Data Location: 16928 - value: (2, 2, 2)
tag: ModelPixelScaleTag (33550) - type: double (12) Tag Location: 166 - Data Location: 16934 - value: (0.25, 0.25, 0.0)
tag: ModelTiepointTag (33922) - type: double (12) Tag Location: 178 - Data Location: 16958 - value: <table: 48 bytes>
tag: GeoKeyDirectoryTag (34735) - type: short (3) Tag Location: 190 - Data Location: 17006 - value: <table: 72 bytes>
tag: GeoDoubleParamsTag (34736) - type: double (12) Tag Location: 202 - Data Location: 17078 - value: <table: 56 bytes>
tag: GeoAsciiParamsTag (34737) - type: string (2) Tag Location: 214 - Data Location: 17134 - value: Amersfoort / RD New|Amersfoort|
tag: ImageWidth (256) - type: short (3) - value: 2924
tag: ImageLength (257) - type: short (3) - value: 2088
tag: BitsPerSample (258) - type: short (3) Tag Location: 46 - Data Location: 218 - value: (32, 32, 32)
tag: Compression (259) - type: short (3) - value: 1
tag: PhotometricInterpretation (262) - type: short (3) - value: 1
tag: StripOffsets (273) - type: long (4) Tag Location: 82 - Data Location: 8576 - value: <table: 8352 bytes>
tag: SamplesPerPixel (277) - type: short (3) - value: 3
tag: RowsPerStrip (278) - type: short (3) - value: 1
tag: StripByteCounts (279) - type: long (4) Tag Location: 118 - Data Location: 224 - value: <table: 8352 bytes>
tag: PlanarConfiguration (284) - type: short (3) - value: 1
tag: ExtraSamples (338) - type: short (3) - value: (0, 0)
tag: SampleFormat (339) - type: short (3) Tag Location: 154 - Data Location: 16928 - value: (2, 2, 2)
tag: ModelPixelScaleTag (33550) - type: double (12) Tag Location: 166 - Data Location: 16934 - value: (0.25, 0.25, 0.0)
tag: ModelTiepointTag (33922) - type: double (12) Tag Location: 178 - Data Location: 16958 - value: <table: 48 bytes>
tag: GeoKeyDirectoryTag (34735) - type: short (3) Tag Location: 190 - Data Location: 17006 - value: <table: 72 bytes>
tag: GeoDoubleParamsTag (34736) - type: double (12) Tag Location: 202 - Data Location: 17078 - value: <table: 56 bytes>
tag: GeoAsciiParamsTag (34737) - type: string (2) Tag Location: 214 - Data Location: 17134 - value: Amersfoort / RD New|Amersfoort|
*** Summary ***
- compression: raw
- photometric_interpretation: 1
- planar_configuration: 1
- fill_order: 1
- size: (2924, 2088)
format key: (b'II', 1, (2, 2, 2), 1, (32, 32, 32), (0, 0))
- unsupported format
Traceback (most recent call last):
File "/Users/mjpieters/Development/venvs/Whosebug-3.6/lib/python3.6/site-packages/PIL/TiffImagePlugin.py", line 1196, in _setup
self.mode, rawmode = OPEN_INFO[key]
KeyError: (b'II', 1, (2, 2, 2), 1, (32, 32, 32), (0, 0))
During handling of the above exception, another exception occurred:
Traceback (most recent call last):
File "<stdin>", line 1, in <module>
File "/Users/mjpieters/Development/venvs/Whosebug-3.6/lib/python3.6/site-packages/PIL/ImageFile.py", line 102, in __init__
self._open()
File "/Users/mjpieters/Development/venvs/Whosebug-3.6/lib/python3.6/site-packages/PIL/TiffImagePlugin.py", line 950, in _open
self._seek(0)
File "/Users/mjpieters/Development/venvs/Whosebug-3.6/lib/python3.6/site-packages/PIL/TiffImagePlugin.py", line 1017, in _seek
self._setup()
File "/Users/mjpieters/Development/venvs/Whosebug-3.6/lib/python3.6/site-packages/PIL/TiffImagePlugin.py", line 1200, in _setup
raise SyntaxError("unknown pixel mode")
SyntaxError: unknown pixel mode
我将 tiff 图像文件从 64 位格式转换为 32 位格式并且成功了。
您可以使用 ImageMagick 7 做到这一点：
magick oldimage.tiff -depth 8 newimage.tif
或使用 ImageMagick 6 或更早版本：
convert out_0.tiff -depth 8 newimage.tiff
pyvips 可以像这样直接使用 TIFF。它将您的文件加载为具有两个 alpha 通道的单声道 int32 图像。
$ python3
Python 3.8.2 (default, Apr 27 2020, 15:53:34)
[GCC 9.3.0] on linux
Type "help", "copyright", "credits" or "license" for more information.
>>> import pyvips
>>> x = pyvips.Image.new_from_file("rgb_CGI.tiff")
>>> x.width, x.height, x.bands, x.format, x.interpretation
(2924, 2088, 3, 'int', 'b-w')
>>> # getpoint reads out a pixel
>>> x.getpoint(10, 10)
[156.0, 141.0, 133.0]
>>> # avg finds the image average
>>> x.avg()
125.31475912560515
>>>
它忽略了额外的 GDAL 标签，但您可以进行所有常见的 2D 图像处理，而且它是 quite a bit quicker than PIL。
我正在尝试使用 PIL 的图像功能打开一个 geotiff 文件。它引发错误：
OSError: cannot identify image file 'Whatever\image\I\use.tiff'
我看到问题被问到 here 例如，解决方案要么使用
Import Image
而不是
From PIL import Image
我认为这是一个过时的解决方案；我无法导入图像。另一个解决方案是将 pillow 更新到 2.9，但 5 年后我们使用的是 5.0.0。我也尝试了 4.0.0，但我收到了同样的错误。 是否有最新的解决方案？
这是我的代码，here 是一个 link 文件：
image_path = 'each\image\I\use.tiff'
from PIL import Image
Image.open(image_path)
您有一个 multi-channel 32 位 TIFF 图像，Pillow 尚不支持该格式。见 issue #1888:
Pillow (and PIL) is currently able to open 8 bit per channel multi-channel images (such as RGB) but is able to open higher bit depth images (e.g. I16, I32, or Float32 images) if they are single channel (e.g., grayscale).
[...]
Requirements
- We should be able to support common GIS formats as well as high bit depth RGB(A) images.
- At least 4 channels, but potentially more (see #1839)
- Different pixel formats, including I16, I32, and Float.
我通过直接使用 TIFF 插件图像 reader 确定了这一点，并启用了调试模式：
from PIL import TiffImagePlugin
TiffImagePlugin.DEBUG = True
with open(image_path, 'rb') as f:
TiffImagePlugin.TiffImageFile(f)
其中包括输出：
tag: BitsPerSample (258) - type: short (3) Tag Location: 46 - Data Location: 218 - value: (32, 32, 32)
（下面是完整的调试输出）
您可以使用 Python GDAL bindings to read this format. You can also use the gdal_translate command line utility 将您的文件转换为 Pillow 可以 处理的格式；对于多波段，您必须降低到 8 位，或者移动到灰度。
例如，要将您的输入文件转换为 PNG，您可以使用：
gdal_translate -of PNG rgb_CGI.tiff rgb_CGI.png
之后 Pillow 可以打开 PNG 文件。
Pillow 的 TIFF 插件的完整调试输出：
>>> from PIL import TiffImagePlugin
>>> TiffImagePlugin.DEBUG = True
>>> with open(image_path, 'rb') as f:
... TiffImagePlugin.TiffImageFile(f)
...
*** TiffImageFile._open ***
- __first: 8
- ifh: b'II*\x00\x08\x00\x00\x00'
Seeking to frame 0, on frame -1, __next 8, location: 8
Loading tags, location: 8
tag: ImageWidth (256) - type: short (3) - value: 2924
tag: ImageLength (257) - type: short (3) - value: 2088
tag: BitsPerSample (258) - type: short (3) Tag Location: 46 - Data Location: 218 - value: (32, 32, 32)
tag: Compression (259) - type: short (3) - value: 1
tag: PhotometricInterpretation (262) - type: short (3) - value: 1
tag: StripOffsets (273) - type: long (4) Tag Location: 82 - Data Location: 8576 - value: <table: 8352 bytes>
tag: SamplesPerPixel (277) - type: short (3) - value: 3
tag: RowsPerStrip (278) - type: short (3) - value: 1
tag: StripByteCounts (279) - type: long (4) Tag Location: 118 - Data Location: 224 - value: <table: 8352 bytes>
tag: PlanarConfiguration (284) - type: short (3) - value: 1
tag: ExtraSamples (338) - type: short (3) - value: (0, 0)
tag: SampleFormat (339) - type: short (3) Tag Location: 154 - Data Location: 16928 - value: (2, 2, 2)
tag: ModelPixelScaleTag (33550) - type: double (12) Tag Location: 166 - Data Location: 16934 - value: (0.25, 0.25, 0.0)
tag: ModelTiepointTag (33922) - type: double (12) Tag Location: 178 - Data Location: 16958 - value: <table: 48 bytes>
tag: GeoKeyDirectoryTag (34735) - type: short (3) Tag Location: 190 - Data Location: 17006 - value: <table: 72 bytes>
tag: GeoDoubleParamsTag (34736) - type: double (12) Tag Location: 202 - Data Location: 17078 - value: <table: 56 bytes>
tag: GeoAsciiParamsTag (34737) - type: string (2) Tag Location: 214 - Data Location: 17134 - value: Amersfoort / RD New|Amersfoort|
tag: ImageWidth (256) - type: short (3) - value: 2924
tag: ImageLength (257) - type: short (3) - value: 2088
tag: BitsPerSample (258) - type: short (3) Tag Location: 46 - Data Location: 218 - value: (32, 32, 32)
tag: Compression (259) - type: short (3) - value: 1
tag: PhotometricInterpretation (262) - type: short (3) - value: 1
tag: StripOffsets (273) - type: long (4) Tag Location: 82 - Data Location: 8576 - value: <table: 8352 bytes>
tag: SamplesPerPixel (277) - type: short (3) - value: 3
tag: RowsPerStrip (278) - type: short (3) - value: 1
tag: StripByteCounts (279) - type: long (4) Tag Location: 118 - Data Location: 224 - value: <table: 8352 bytes>
tag: PlanarConfiguration (284) - type: short (3) - value: 1
tag: ExtraSamples (338) - type: short (3) - value: (0, 0)
tag: SampleFormat (339) - type: short (3) Tag Location: 154 - Data Location: 16928 - value: (2, 2, 2)
tag: ModelPixelScaleTag (33550) - type: double (12) Tag Location: 166 - Data Location: 16934 - value: (0.25, 0.25, 0.0)
tag: ModelTiepointTag (33922) - type: double (12) Tag Location: 178 - Data Location: 16958 - value: <table: 48 bytes>
tag: GeoKeyDirectoryTag (34735) - type: short (3) Tag Location: 190 - Data Location: 17006 - value: <table: 72 bytes>
tag: GeoDoubleParamsTag (34736) - type: double (12) Tag Location: 202 - Data Location: 17078 - value: <table: 56 bytes>
tag: GeoAsciiParamsTag (34737) - type: string (2) Tag Location: 214 - Data Location: 17134 - value: Amersfoort / RD New|Amersfoort|
*** Summary ***
- compression: raw
- photometric_interpretation: 1
- planar_configuration: 1
- fill_order: 1
- size: (2924, 2088)
format key: (b'II', 1, (2, 2, 2), 1, (32, 32, 32), (0, 0))
- unsupported format
Traceback (most recent call last):
File "/Users/mjpieters/Development/venvs/Whosebug-3.6/lib/python3.6/site-packages/PIL/TiffImagePlugin.py", line 1196, in _setup
self.mode, rawmode = OPEN_INFO[key]
KeyError: (b'II', 1, (2, 2, 2), 1, (32, 32, 32), (0, 0))
During handling of the above exception, another exception occurred:
Traceback (most recent call last):
File "<stdin>", line 1, in <module>
File "/Users/mjpieters/Development/venvs/Whosebug-3.6/lib/python3.6/site-packages/PIL/ImageFile.py", line 102, in __init__
self._open()
File "/Users/mjpieters/Development/venvs/Whosebug-3.6/lib/python3.6/site-packages/PIL/TiffImagePlugin.py", line 950, in _open
self._seek(0)
File "/Users/mjpieters/Development/venvs/Whosebug-3.6/lib/python3.6/site-packages/PIL/TiffImagePlugin.py", line 1017, in _seek
self._setup()
File "/Users/mjpieters/Development/venvs/Whosebug-3.6/lib/python3.6/site-packages/PIL/TiffImagePlugin.py", line 1200, in _setup
raise SyntaxError("unknown pixel mode")
SyntaxError: unknown pixel mode
我将 tiff 图像文件从 64 位格式转换为 32 位格式并且成功了。
您可以使用 ImageMagick 7 做到这一点：
magick oldimage.tiff -depth 8 newimage.tif
或使用 ImageMagick 6 或更早版本：
convert out_0.tiff -depth 8 newimage.tiff
pyvips 可以像这样直接使用 TIFF。它将您的文件加载为具有两个 alpha 通道的单声道 int32 图像。
$ python3
Python 3.8.2 (default, Apr 27 2020, 15:53:34)
[GCC 9.3.0] on linux
Type "help", "copyright", "credits" or "license" for more information.
>>> import pyvips
>>> x = pyvips.Image.new_from_file("rgb_CGI.tiff")
>>> x.width, x.height, x.bands, x.format, x.interpretation
(2924, 2088, 3, 'int', 'b-w')
>>> # getpoint reads out a pixel
>>> x.getpoint(10, 10)
[156.0, 141.0, 133.0]
>>> # avg finds the image average
>>> x.avg()
125.31475912560515
>>>
它忽略了额外的 GDAL 标签，但您可以进行所有常见的 2D 图像处理，而且它是 quite a bit quicker than PIL。