ValueError：参数必须是密集张量–Python和TensorFlow

我正在提取可能与我遇到的问题相关的代码的一些部分：

from PIL import Image
import tensorflow as tf

data = Image.open('1-enhanced.png')
...
...
raw_data = data
raw_img = raw_data

我收到以下长消息,我不确定如何分析(你对这里发生了什么有任何想法)：

Traceback (most recent call last):
  File "C:Python35libsite-packagestensorflowpythonframeworkop_def_library.py",line 490,in apply_op
    preferred_dtype=default_dtype)
  File "C:Python35libsite-packagestensorflowpythonframeworkops.py",line 669,in convert_to_tensor
    ret = conversion_func(value,dtype=dtype,name=name,as_ref=as_ref)
  File "C:Python35libsite-packagestensorflowpythonframeworkconstant_op.py",line 176,in _constant_tensor_conversion_function
    return constant(v,name=name)
  File "C:Python35libsite-packagestensorflowpythonframeworkconstant_op.py",line 165,in constant
    tensor_util.make_tensor_proto(value,shape=shape,verify_shape=verify_shape))
  File "C:Python35libsite-packagestensorflowpythonframeworktensor_util.py",line 376,in make_tensor_proto
    _GetDenseDimensions(values)))
ValueError: Argument must be a dense tensor: <PIL.PngImagePlugin.PngImageFile image mode=L size=150x150 at 0x1E07D3C0AC8> - got shape [150,150],but wanted [].

During handling of the above exception,another exception occurred:

Traceback (most recent call last):
  File "conv_visuals.py",line 54,in <module>
    x = tf.reshape(raw_data,shape=[-1,150,1])
  File "C:Python35libsite-packagestensorflowpythonopsgen_array_ops.py",line 2448,in reshape
    name=name)
  File "C:Python35libsite-packagestensorflowpythonframeworkop_def_library.py",line 503,in apply_op
    as_ref=input_arg.is_ref).dtype.name
  File "C:Python35libsite-packagestensorflowpythonframeworkops.py",but wanted [].

谢谢.

解决方法

尝试将其转换为numpy数组：

numpy.asarray(Image.open('1-enhanced.png').convert('L'))

（编辑：台州站长网）

【声明】本站内容均来自网络，其相关言论仅代表作者个人观点，不代表本站立场。若无意侵犯到您的权利，请及时与联系站长删除相关内容!