检查时出错：预期 dense_3_input 有 2 维，但得到的数组形状为 (28, 28, 1)

Question

我正在写一个神经元来确定手写数字

import numpy as np
from keras.utils import np_utils
from keras.models import model_from_json
from keras.preprocessing import image
import matplotlib.pyplot as plt

json_file = open("mnist_model.json", "r")
loaded_model_json = json_file.read()
json_file.close()
loaded_model = model_from_json(loaded_model_json)
loaded_model.load_weights("mnist_model.h5")


loaded_model.compile(loss= "categorical_crossentropy", optimizer="adam", metrics=["accuracy"])


img_path ="5.png"
img = image.load_img(img_path, target_size=(28,28), grayscale=True)
plt.imshow(img, cmap='gray')
plt.show

x =image.img_to_array(img)
x = 255 - x 
x/= 255
np.expand_dims(x, axis=0)
prediction = loaded_model.predict(x)
prediction = np_utils.categorical_pobabs_to_classes(prediction)
print(prediction)

我所做的只是教她使用它，但随后问题就出现了： 1.结果是一个图表和错误ValueError: Error when checking : expected dense_3_input to have 2 dimensions, but got array with shape (28, 28, 1) in Line ´img = image.load_img(img_path, target_size=(28,28), grayscale=True)´

Answer 1

我认为你的错在这一行

np.expand_dims(x, axis=0)

应该是：

x = np.expand_dims(x, axis=0)