這是我第一次創建自定義數據生成器。我對自定義數據生成器還不是很熟悉
為什麼 np.load 無法加載 .npy 文件？我在創建 npy 文件時遇到錯誤嗎？
data/syu/npy 中的所有 image_folders 和 label_folders，我嘗試使用與 ImageDataGenerator 相同的文件夾結構。
我的圖像是 5 * 1 通道和 1 * 3 通道，我concatenate到 8 通道，但單通道影像不能concatenate，所以我將它們擴展到 3 通道並concatenate成 18 通道 .npy 文件，label也執行相同的操作。
TypeError
expected str, bytes or os.PathLike object, not numpy.int32
File "C:\Labbb\testing\unet_mao0\data.py", line 219, in get_data
image = np.load(image_path)
File "C:\Labbb\testing\unet_mao0\data.py", line 205, in __getitem__
return self.get_data(batch)
File "C:\Labbb\testing\unet_mao0\main.py", line 105, in <module>
train_history=model.fit_generator(data_generator, steps_per_epoch=200,epochs=30,callbacks=[model_checkpoint])
TypeError: expected str, bytes or os.PathLike object, not numpy.int32
main.py
params = {'dim': (512,512), 'batch_size': 1, 'n_classes': 7, 'n_channels': 18, 'shuffle': True}
image_folders = ['image0', 'image1', 'image2', 'image3', 'image4', 'image6', 'image7']
label_folders = ['label0', 'label1', 'label2', 'label3', 'label4', 'label6', 'label7']
data_generator = CustomDataGenerator(image_folders, label_folders, **params)
model = unet()
model_checkpoint = ModelCheckpoint('weight/syu/0123467_Custom_test1.hdf5', monitor='loss',verbose=1, save_best_only=True)
class CustomDataGenerator(Sequence):
def __init__(self, image_folders, label_folders,dim=(512,512), batch_size=1,n_classes=7,n_channels=18,shuffle=True):
self.image_folders = image_folders
self.label_folders = label_folders
self.dim = dim
self.batch_size = batch_size
self.n_classes = n_classes
self.n_channels = n_channels
self.shuffle = shuffle
self.image_paths = []
self.label_paths = []
for folder in self.image_folders:
image_folder_path = os.path.join('data/syu/npy', folder)
image_files = os.listdir(image_folder_path)
for file_name in image_files:
self.image_paths.append(os.path.join(image_folder_path, file_name))
for folder in self.label_folders:
label_folder_path = os.path.join('data/syu/npy', folder)
label_files = os.listdir(label_folder_path)
for file_name in label_files:
self.label_paths.append(os.path.join(label_folder_path, file_name))
self.on_epoch_end()
def __len__(self):
return int(np.ceil(len(self.image_paths) / self.batch_size))
def __getitem__(self, index):
batch_image_paths = self.image_paths[index * self.batch_size: (index + 1) * self.batch_size]
batch_label_paths = self.label_paths[index * self.batch_size: (index + 1) * self.batch_size]
batch = zip(batch_image_paths, batch_label_paths)
return self.get_data(batch)
def on_epoch_end(self):
self.image_paths = np.arange(len(self.image_paths))
self.label_paths = np.arange(len(self.label_paths))
if self.shuffle == True:
np.random.shuffle(self.image_paths)
np.random.shuffle(self.label_paths)
def get_data(self, batch):
X = np.empty((self.batch_size, *self.dim, self.n_channels))
y = np.empty((self.batch_size, *self.dim, self.n_classes))
for i, (image_path, label_path) in enumerate(batch):
image = np.load(image_path)
label = np.load(label_path)
X[i,] = image
y[i,] = label
return X, y
根据错误信息TypeError: expected str, bytes or os.PathLike object, not numpy.int32,问题出在np.load()这里,你传给它的image_path和label_path应该是一个字符串类型的文件路径,但现在是一个numpy整数数组。
主要原因是在on_epoch_end()方法里,你把image_paths和label_paths重新赋值为:
python
self.image_paths = np.arange(len(self.image_paths))
self.label_paths = np.arange(len(self.label_paths))
这将它们变成了整数数组。
你应该删除或者注释掉这两行代码,保持image_paths和label_paths为字符串类型的文件路径数组,不要重新赋值成整数数组。
另外一些建议:
__init__里image_paths和label_paths赋值,直接用np.random.shuffle对它们原地shuffle即可。yield代替__getitem__和get_data,简化代码逻辑