tf.io.extract_jpeg_shape

Extract the shape information of a JPEG-encoded image.

View aliases

Main aliases

tf.image.extract_jpeg_shape

Compat aliases for migration

See Migration guide for more details.

tf.compat.v1.image.extract_jpeg_shape, tf.compat.v1.io.extract_jpeg_shape

tf.io.extract_jpeg_shape(
    contents: Annotated[Any, _atypes.String],
    output_type: TV_ExtractJpegShape_output_type = tf.dtypes.int32,
    name=None
) -> Annotated[Any, TV_ExtractJpegShape_output_type]

This op only parses the image header, so it is much faster than DecodeJpeg.

Args
contents	A Tensor of type string. 0-D. The JPEG-encoded image.
output_type	An optional tf.DType from: tf.int32, tf.int64. Defaults to tf.int32. (Optional) The output type of the operation (int32 or int64). Defaults to int32.
name	A name for the operation (optional).

Returns
A Tensor of type output_type.