#image_batch = "malign.png"
classes = os.listdir("/content/drive/MyDrive/DDSM Dataset/")
matrix = {
"Benign Masses1":0, # bon
"Malignant Masses1": 0, # mauvais
"Benign Masses2":0, # mauvais
"Malignant Masses2": 0 # bon
}
i, s = 0, 0
for classe in classes:
i+=1
path = f"drive/MyDrive/MIAS_Dataset/{classe}/"
images = os.listdir(path)
for image in images:
img = tf.keras.utils.load_img(path+image,
target_size=(img_height, img_width))
img_array = tf.keras.utils.img_to_array(img)
img_array = tf.expand_dims(img_array, 0) # Create a batch
predictions = model.predict(img_array)
score = tf.nn.softmax(predictions[0])
if class_names[np.argmax(score)] == classe:
matrix[f"{classe}{i}"] += 1 # if the class is correct
else:
if i == 1 :
matrix[f"Benign Masses{2}"] += 1 # if the class is not correct (i==1), then save in the other classe (i==2)
else:
matrix[f"Malignant Masses{1}"] += 1 # if the class is not correct (i==2), then save in the other classe (i==1)
if s%100 == 0:
print(f"{round(s*100/3816)}%")
s +=1
#print(
# "This image most likely belongs to {} with a {:.2f} percent confidence."
# .format(class_names[np.argmax(score)], 100 * np.max(score)))