ValueError：检查输入时出错：预期 input_1 的形状为 (224, 224, 3) 但得到的数组的形状为 (3, 224, 224)

Question

我在视频数据集上使用预训练模型。下面是我的代码。

  from keras.applications.vgg19 import decode_predictions
  from keras.applications.vgg19 import VGG19, preprocess_input
  import threading,cv2
  import numpy as np
  label = ''
  frame = None
  class MyThread(threading.Thread):
  def __init__(self):
    threading.Thread.__init__(self)
  def run(self):
    global label
    self.model = VGG19(weights="imagenet")
    while (~(frame is None)):
        (inID, label) = self.predict(frame)
  def predict(self, frame):
    image = cv2.cvtColor(frame, cv2.COLOR_BGR2RGB).astype(np.float32)
    image = image.transpose((2, 0, 1))
    image = image.reshape((1,) + image.shape)
    image = preprocess_input(image)
    preds = self.model.predict(image)
    return decode_predictions(preds)[0]
 videoFile ="D:/lostpanda.mp4"
 cap = cv2.VideoCapture(videoFile)
 while(cap.isOpened()):
 keras_thread = MyThread()
 keras_thread.start()
 while (True):
    ret, original = cap.read()
    frame = cv2.resize(original, (224, 224))
    cv2.putText(original, "Label: {}".format(label), (10, 30), cv2.FONT_HERSHEY_SIMPLEX, 0.9, (0, 255, 0), 2)
    cv2.imshow("Classification", original)
cap.release()
frame = None

我正在检索以下错误

  ValueError: Error when checking input: expected input_1 to have a shape (224, 224, 3) but got array 
  with shape (3, 224, 224)
  Traceback (most recent call last):
  File "C:/Usersvideo-classification.py", line 35, in <module>
      frame = cv2.resize(original, (224, 224))
    cv2.error: OpenCV(4.2.0) C:\projects\opencv-python\opencv\modules\imgproc\src\resize.cpp:4045: error: 
    (-215:Assertion failed) !ssize.empty() in function 'cv::resize'

谢谢，非常感谢您的帮助。

Answer 1

老实说，为了能够在视频上显示标签，我们需要区分两个问题：

1. 准备输入形状和类型以匹配模型要求。
2. 使用线程来提高性能。

在下面的代码中预测有效，因此您可以将其用作起点。之后，您将需要引入线程以提高性能并加快预测速度。

from keras.applications.vgg19 import VGG19, preprocess_input
from keras.applications.vgg19 import decode_predictions
from keras.preprocessing.image import img_to_array

import threading
import cv2
import numpy as np

def predict_custom(image):

  # convert the image pixels to a numpy array
  image = img_to_array(image)

  # reshape data for the model
  image = image.reshape((1, image.shape[0], image.shape[1], image.shape[2]))
  
  # prepare the image for the VGG model
  image = preprocess_input(image)
   
  # predict the probability across all output classes
  preds = model.predict(image)
  return decode_predictions(preds)[0]

# define the model
model = VGG19(weights="imagenet")

# Load the video file
videoFile ="file_example_MP4_480_1_5MG.mp4"
cap = cv2.VideoCapture(videoFile)
(W, H) = (None, None)
writer = None

# loop over frames from the video file stream
while True:
    # read the next frame from the file
    (grabbed, frame) = cap.read()
    
    # the stream ends when the frame was not grabbed
    if not grabbed:
        break
    
    # if the frame dimensions are empty, grab them
    if W is None or H is None:
        (H, W) = frame.shape[:2]

    # clone frame to use it for output later
    output = frame.copy()

    # convert the frame into gray
    frame = cv2.cvtColor(frame, cv2.COLOR_BGR2RGB)
    
    # resize the frame to a fixed 224x224
    frame = cv2.resize(frame, (224, 224)).astype("float32")
      
    # perfrom the  prediction
    label = predict_custom(frame)
    
    # show the output image
    cv2.putText(output, "Label: {}".format(label), (10, 30), cv2.FONT_HERSHEY_SIMPLEX, 0.9, (0, 255, 0), 2)
    cv2.imshow("Classification", output)
    key = cv2.waitKey(1) & 0xFF

cap.release()
frame = None