将图像另存为numpy数组

Question

I am not able to load images into numpy array and getting an error like this...

ValueError: could not broadcast input array from shape (175,217,3) into shape (100,100,3)

The function code:

import cv2
import numpy as np
import os

train_data_dir = '/home/ec2-user/SageMaker/malaria-detection-model/malaria/training'
valid_data_dir = '/home/ec2-user/SageMaker/malaria-detection-model/malaria/validation'

# declare the number of samples in each category
nb_train_samples = 22045 #  training samples
nb_valid_samples = 5513#  validation samples
num_classes = 2
img_rows_orig = 100
img_cols_orig = 100

def load_training_data():
    labels = os.listdir(train_data_dir)
    total = len(labels)

    X_train = np.ndarray((nb_train_samples, img_rows_orig, img_cols_orig, 3), dtype=np.uint8)
    Y_train = np.zeros((nb_train_samples,), dtype='uint8')

    i = 0
    j = 0
    for label in labels:
        image_names_train = os.listdir(os.path.join(train_data_dir, label))
        total = len(image_names_train)
        print(label, total)
        for image_name in image_names_train:
            img = cv2.imread(os.path.join(train_data_dir, label, image_name), cv2.IMREAD_COLOR)
            img = np.array([img])
            X_train[i] = img
            Y_train[i] = j

            if i % 100 == 0:
                print('Done: {0}/{1} images'.format(i, total))
            i += 1
        j += 1    
    print(i)                
    print('Loading done.')

    np.save('imgs_train.npy', X_train, Y_train)
    return X_train, Y_train

This function is part of the file load_data.py that can be found in malaria_cell_classification_code.zip file from:

https://ceb.nlm.nih.gov/repositories/malaria-datasets/

---

I tried to change X_train and Y_train to list instead of numpy array. The function halts at np.save method.

X_train = Y_train = list()
        X_train.append(img)
        Y_train.append(j)

What is the correct and standard way to save images in numpy?

---

After resizing the image, I get different error:

Done: 19400/9887 images
Done: 19500/9887 images
Done: 19600/9887 images
Done: 19700/9887 images
Done: 19800/9887 images
19842
Loading done.
Transform targets to keras compatible format.
Done: 19800/9887 images
19842
Loading done.
Transform targets to keras compatible format.
------------------------------
Creating validation images...
------------------------------
Parasitized 1098
---------------------------------------------------------------------------
error                                     Traceback (most recent call last)
<ipython-input-6-8008be74f482> in <module>()
      2 #load data for training
      3 X_train, Y_train = load_resized_training_data(img_rows, img_cols)
----> 4 X_valid, Y_valid = load_resized_validation_data(img_rows, img_cols)
      5 #print the shape of the data
      6 print(X_train.shape, Y_train.shape, X_valid.shape, Y_valid.shape)

~/SageMaker/malaria-detection-model/malaria_cell_classification_code/load_data.py in load_resized_validation_data(img_rows, img_cols)
    103 def load_resized_validation_data(img_rows, img_cols):
    104 
--> 105     X_valid, Y_valid = load_validation_data()
    106 
    107     # Resize images

~/SageMaker/malaria-detection-model/malaria_cell_classification_code/load_data.py in load_validation_data()
     75 
     76             img = np.array([img])
---> 77             img2 = cv2.resize(img, (100, 100))
     78             X_valid[i] = img2
     79             Y_valid[i] = j

error: OpenCV(4.0.0) /io/opencv/modules/imgproc/src/resize.cpp:3427: error: (-215:Assertion failed) !dsize.empty() in function 'resize'


------------------------------
Creating validation images...
------------------------------
Parasitized 1098
---------------------------------------------------------------------------
error                                     Traceback (most recent call last)
<ipython-input-6-8008be74f482> in <module>()
      2 #load data for training
      3 X_train, Y_train = load_resized_training_data(img_rows, img_cols)
----> 4 X_valid, Y_valid = load_resized_validation_data(img_rows, img_cols)
      5 #print the shape of the data
      6 print(X_train.shape, Y_train.shape, X_valid.shape, Y_valid.shape)

~/SageMaker/malaria-detection-model/malaria_cell_classification_code/load_data.py in load_resized_validation_data(img_rows, img_cols)
    103 def load_resized_validation_data(img_rows, img_cols):
    104 
--> 105     X_valid, Y_valid = load_validation_data()
    106 
    107     # Resize images

~/SageMaker/malaria-detection-model/malaria_cell_classification_code/load_data.py in load_validation_data()
     75 
     76             img = np.array([img])
---> 77             img2 = cv2.resize(img, (100, 100))
     78             X_valid[i] = img2
     79             Y_valid[i] = j

error: OpenCV(4.0.0) /io/opencv/modules/imgproc/src/resize.cpp:3427: error: (-215:Assertion failed) !dsize.empty() in function 'resize'

The complete script can be found here...

https://gist.github.com/shantanuo/cfe0913b367647890451f5ae3f6fb691

Answer 1

opencv2 already returns a numpy array. Don't make a new one, especially not one with an additional level of nesting:

img = cv2.imread(os.path.join(train_data_dir, label, image_name), cv2.IMREAD_COLOR)
img = cv2.resize(img, (100, 100))