tf.keras.applications.vgg16.preprocess_input

Preprocesses a tensor or Numpy array encoding a batch of images.

tf.keras.applications.vgg16.preprocess_input(
    x, data_format=None
)

Usage example with applications.MobileNet:

i = keras.layers.Input([None, None, 3], dtype="uint8")
x = ops.cast(i, "float32")
x = keras.applications.mobilenet.preprocess_input(x)
core = keras.applications.MobileNet()
x = core(x)
model = keras.Model(inputs=[i], outputs=[x])
result = model(image)

x A floating point numpy.array or a backend-native tensor, 3D or 4D with 3 color channels, with values in the range [0, 255]. The preprocessed data are written over the input data if the data types are compatible. To avoid this behaviour, numpy.copy(x) can be used.
data_format Optional data format of the image tensor/array. None, means the global setting keras.backend.image_data_format() is used (unless you changed it, it uses "channels_last"). Defaults to None.

Preprocessed array with type float32.

The images are converted from RGB to BGR, then each color channel is zero-centered with respect to the ImageNet dataset, without scaling.

ValueError In case of unknown data_format argument.