cutcutcodec.core.io.read_image

Read an image with opencv.

Classes

ContainerInputImage(filename)

Decode an image.

Functions

read_image(filename)

Read the image and make it compatible with Video Frame.

Details

class cutcutcodec.core.io.read_image.ContainerInputImage(filename: Path | str | bytes)[source]

Decode an image.

Attributes

filenamepathlib.Path: The path to the physical file that contains the extracted image stream (readonly).

Examples

>>> from cutcutcodec.core.io.read_image import ContainerInputImage
>>> from cutcutcodec.utils import get_project_root
>>> (stream,) = ContainerInputImage(get_project_root() / "examples" / "logo.png").out_streams
>>> stream.snapshot(0, (9, 9))[..., 3]  # alpha layer
tensor([[0.0000, 0.0415, 0.5152, 0.8748, 0.9872, 0.8744, 0.5164, 0.0422, 0.0000],
        [0.0418, 0.7853, 1.0000, 1.0000, 1.0000, 1.0000, 1.0000, 0.7851, 0.0420],
        [0.5156, 1.0000, 1.0000, 1.0000, 1.0000, 1.0000, 1.0000, 1.0000, 0.5141],
        [0.8749, 1.0000, 1.0000, 1.0000, 1.0000, 1.0000, 1.0000, 1.0000, 0.8732],
        [0.9871, 1.0000, 1.0000, 1.0000, 1.0000, 1.0000, 1.0000, 1.0000, 0.9861],
        [0.8745, 1.0000, 1.0000, 1.0000, 1.0000, 1.0000, 1.0000, 1.0000, 0.8727],
        [0.5150, 1.0000, 1.0000, 1.0000, 1.0000, 1.0000, 1.0000, 1.0000, 0.5137],
        [0.0417, 0.7838, 1.0000, 1.0000, 1.0000, 1.0000, 1.0000, 0.7842, 0.0413],
        [0.0000, 0.0411, 0.5139, 0.8732, 0.9865, 0.8729, 0.5144, 0.0417, 0.0000]])
>>>

Initialise and create the class.

Parameters

filenamepathlike: Path to the file to be decoded.

Raises

cutcutcodec.core.exceptions.DecodeError: If it fail to extract any multimedia stream from the provided file.

cutcutcodec.core.io.read_image.read_image(filename: Path | str | bytes) → tuple[Tensor, Colorspace][source]

Read the image and make it compatible with Video Frame.

Parameters

filenamepathlike: The pathlike of the image file.

Returns

imagetorch.Tensor: The image in float32 of shape (height, width, channels).
colorspaceColorspace: The color space in which the image is defined. No conversion performed.

Raises

cutcutcodec.core.exceptions.DecodeError: If it fails to read the image.

Notes

Does not care about colorspace, no conversions performed.

Examples

>>> from cutcutcodec.core.io.read_image import read_image
>>> from cutcutcodec.utils import get_project_root
>>> for file in sorted((get_project_root().parent / "media" / "image").glob("image.*")):
...     image, colorspace = read_image(file)
...     print(f"{file.name}: {image.shape}, {colorspace}")
...
image.avif: torch.Size([64, 64, 3]), Colorspace("y'pbpr", 'bt709', 'iec61966-2-1, iec61966_2_1')
image.bmp: torch.Size([64, 64, 3]), Colorspace("r'g'b'", 'bt709', 'iec61966-2-1, iec61966_2_1')
image.exr: torch.Size([64, 64, 3]), Colorspace("r'g'b'", 'bt709', 'linear')
image.heic: torch.Size([64, 64, 3]), Colorspace("y'pbpr", 'bt709', 'iec61966-2-1, iec61966_2_1')
image.jp2: torch.Size([64, 64, 3]), Colorspace("r'g'b'", 'bt709', 'iec61966-2-1, iec61966_2_1')
image.jpg: torch.Size([64, 64, 3]), Colorspace("y'pbpr", 'bt709', 'iec61966-2-1, iec61966_2_1')
image.kra: torch.Size([64, 64, 3]), Colorspace("r'g'b'", 'bt709', 'iec61966-2-1, iec61966_2_1')
image.pbm: torch.Size([64, 64, 1]), Colorspace("y'pbpr", 'bt709', 'iec61966-2-1, iec61966_2_1')
image.pgm: torch.Size([64, 64, 1]), Colorspace("y'pbpr", 'bt709', 'iec61966-2-1, iec61966_2_1')
image.png: torch.Size([64, 64, 4]), Colorspace("r'g'b'", 'bt709', 'iec61966-2-1, iec61966_2_1')
image.pnm: torch.Size([64, 64, 3]), Colorspace("r'g'b'", 'bt709', 'iec61966-2-1, iec61966_2_1')
image.ppm: torch.Size([64, 64, 3]), Colorspace("r'g'b'", 'bt709', 'iec61966-2-1, iec61966_2_1')
image.psd: torch.Size([64, 64, 3]), Colorspace("r'g'b'", 'bt709', 'iec61966-2-1, iec61966_2_1')
image.ras: torch.Size([64, 64, 3]), Colorspace("r'g'b'", 'bt709', 'iec61966-2-1, iec61966_2_1')
image.sgi: torch.Size([64, 64, 3]), Colorspace("r'g'b'", 'bt709', 'iec61966-2-1, iec61966_2_1')
image.tiff: torch.Size([64, 64, 3]), Colorspace("r'g'b'", 'bt709', 'iec61966-2-1, iec61966_2_1')
image.webp: torch.Size([64, 64, 3]), Colorspace("y'pbpr", 'bt709', 'iec61966-2-1, iec61966_2_1')
image.xbm: torch.Size([64, 64, 1]), Colorspace("y'pbpr", 'bt709', 'iec61966-2-1, iec61966_2_1')