Missed TensorFlow World? Check out the recap. Learn more

tf.io.extract_jpeg_shape

TensorFlow 2.0 version

Defined in generated file: python/ops/gen_image_ops.py

Extract the shape information of a JPEG-encoded image.

Aliases:

  • tf.compat.v1.image.extract_jpeg_shape
  • tf.compat.v1.io.extract_jpeg_shape
  • tf.compat.v2.image.extract_jpeg_shape
  • tf.compat.v2.io.extract_jpeg_shape
  • tf.image.extract_jpeg_shape
tf.io.extract_jpeg_shape(
    contents,
    output_type=tf.dtypes.int32,
    name=None
)

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.