Show / Hide Table of Contents

Class ImageDataGenerator

Generate batches of tensor image data with real-time data augmentation. The data will be looped over (in batches).

Inheritance
System.Object
Keras
Base
ImageDataGenerator
Implements
System.IDisposable
Inherited Members
Base.Parameters
Base.None
Base.Init()
Base.ToPython()
Base.InvokeStaticMethod(Object, String, Dictionary<String, Object>)
Base.InvokeMethod(String, Dictionary<String, Object>)
Base.Item[String]
Keras.Instance
Keras.keras
Keras.keras2onnx
Keras.tfjs
Keras.Dispose()
Keras.ToTuple(Array)
Keras.ToList(Array)
System.Object.Equals(System.Object)
System.Object.Equals(System.Object, System.Object)
System.Object.GetHashCode()
System.Object.GetType()
System.Object.MemberwiseClone()
System.Object.ReferenceEquals(System.Object, System.Object)
System.Object.ToString()
Namespace: Keras.PreProcessing.Image
Assembly: Keras.dll
Syntax
public class ImageDataGenerator : Base, IDisposable

Constructors

| Improve this Doc View Source

ImageDataGenerator(Boolean, Boolean, Boolean, Boolean, Boolean, Single, Int32, Single, Single, Single[], Single, Single, Single, String, Single, Boolean, Boolean, Nullable<Single>, PyObject, String, Single, String)

Initializes a new instance of the ImageDataGenerator class.

Declaration
public ImageDataGenerator(bool featurewise_center = false, bool samplewise_center = false, bool featurewise_std_normalization = false, bool samplewise_std_normalization = false, bool zca_whitening = false, float zca_epsilon = 1E-06F, int rotation_range = 0, float width_shift_range = 0F, float height_shift_range = 0F, float[] brightness_range = null, float shear_range = 0F, float zoom_range = 0F, float channel_shift_range = 0F, string fill_mode = "nearest", float cval = 0F, bool horizontal_flip = false, bool vertical_flip = false, float? rescale = default(float? ), PyObject preprocessing_function = null, string data_format = "", float validation_split = 0F, string dtype = "")
Parameters
Type Name Description
System.Boolean featurewise_center

Boolean. Set input mean to 0 over the dataset, feature-wise.

System.Boolean samplewise_center

Boolean. Set each sample mean to 0.

System.Boolean featurewise_std_normalization

Boolean. Divide inputs by std of the dataset, feature-wise.

System.Boolean samplewise_std_normalization

Boolean. Divide each input by its std.

System.Boolean zca_whitening

Boolean. Apply ZCA whitening.

System.Single zca_epsilon

epsilon for ZCA whitening. Default is 1e-6.

System.Int32 rotation_range

Int. Degree range for random rotations.

System.Single width_shift_range

Float, 1-D array-like or int

System.Single height_shift_range

Float, 1-D array-like or int

System.Single[] brightness_range

Tuple or list of two floats. Range for picking a brightness shift value from.

System.Single shear_range

Float. Shear Intensity (Shear angle in counter-clockwise direction in degrees)

System.Single zoom_range

Float or [lower, upper]. Range for random zoom. If a float, [lower, upper] = [1-zoom_range, 1+zoom_range].

System.Single channel_shift_range

Float. Range for random channel shifts.

System.String fill_mode

One of {"constant", "nearest", "reflect" or "wrap"}. Default is 'nearest'. Points outside the boundaries of the input are filled according to the given mode, 'constant'-> kkkkkkkk|abcd|kkkkkkkk (cval=k), 'nearest'-> aaaaaaaa|abcd|dddddddd, 'reflect'-> abcddcba|abcd|dcbaabcd, 'wrap'-> abcdabcd|abcd|abcdabcd

System.Single cval

Float or Int. Value used for points outside the boundaries when fill_mode = "constant".

System.Boolean horizontal_flip

Boolean. Randomly flip inputs horizontally.

System.Boolean vertical_flip

Boolean. Randomly flip inputs vertically.

System.Nullable<System.Single> rescale

rescaling factor. Defaults to None. If None or 0, no rescaling is applied, otherwise we multiply the data by the value provided (after applying all other transformations).

Python.Runtime.PyObject preprocessing_function

function that will be implied on each input. The function will run after the image is resized and augmented. The function should take one argument: one image (Numpy tensor with rank 3), and should output a Numpy tensor with the same shape.

System.String data_format

Image data format, either "channels_first" or "channels_last". "channels_last" mode means that the images should have shape (samples, height, width, channels), "channels_first" mode means that the images should have shape (samples, channels, height, width). It defaults to the image_data_format value found in your Keras config file at ~/.keras/keras.json. If you never set it, then it will be "channels_last".

System.Single validation_split

Float. Fraction of images reserved for validation (strictly between 0 and 1).

System.String dtype

Dtype to use for the generated arrays.

Methods

| Improve this Doc View Source

ApplyTransform(NDarray, Dictionary<String, Object>)

Applies the transform.

Declaration
public NDarray ApplyTransform(NDarray x, Dictionary<string, object> transform_parameters)
Parameters
Type Name Description
Numpy.NDarray x

3D tensor, single image.

System.Collections.Generic.Dictionary<System.String, System.Object> transform_parameters

Dictionary with string - parameter pairs describing the transformation. Currently, the following parameters from the dictionary are used:

Returns
Type Description
Numpy.NDarray

A transformed version of the input (same shape).

| Improve this Doc View Source

Fit(NDarray, Boolean, Int32, Nullable<Int32>)

Fits the data generator to some sample data. This computes the internal data stats related to the data-dependent transformations, based on an array of sample data. Only required if featurewise_center or featurewise_std_normalization or zca_whitening are set to True.

Declaration
public void Fit(NDarray x, bool augment = false, int rounds = 1, int? seed = default(int? ))
Parameters
Type Name Description
Numpy.NDarray x

The x.

System.Boolean augment

if set to true [augment].

System.Int32 rounds

The rounds.

System.Nullable<System.Int32> seed

The seed.

| Improve this Doc View Source

Flow(NDarray, NDarray, Int32, Boolean, NDarray, Nullable<Int32>, String, String, String, String)

Flows the specified x.

Declaration
public KerasIterator Flow(NDarray x, NDarray y = null, int batch_size = 32, bool shuffle = true, NDarray sample_weight = null, int? seed = default(int? ), string save_to_dir = "", string save_prefix = "", string save_format = "png", string subset = "")
Parameters
Type Name Description
Numpy.NDarray x

Input data. Numpy array of rank 4 or a tuple. If tuple, the first element should contain the images and the second element another numpy array or a list of numpy arrays that gets passed to the output without any modifications. Can be used to feed the model miscellaneous data along with the images. In case of grayscale data, the channels axis of the image array should have value 1, in case of RGB data, it should have value 3, and in case of RGBA data, it should have value 4.

Numpy.NDarray y

Labels.

System.Int32 batch_size

Int (default: 32).

System.Boolean shuffle

Boolean (default: True).

Numpy.NDarray sample_weight

Sample weights.

System.Nullable<System.Int32> seed

Int (default: None).

System.String save_to_dir

None or str (default: None). This allows you to optionally specify a directory to which to save the augmented pictures being generated (useful for visualizing what you are doing).

System.String save_prefix

Str (default: ''). Prefix to use for filenames of saved pictures (only relevant if save_to_dir is set).

System.String save_format

one of "png", "jpeg" (only relevant if save_to_dir is set). Default: "png".

System.String subset

Subset of data ("training" or "validation") if validation_split is set in ImageDataGenerator.

Returns
Type Description
KerasIterator

An Iterator yielding tuples of (x, y) where x is a numpy array of image data (in the case of a single image input) or a list of numpy arrays (in the case with additional inputs) and y is a numpy array of corresponding labels. If 'sample_weight' is not None, the yielded tuples are of the form (x, y, sample_weight). If y is None, only the numpy array x is returned.

| Improve this Doc View Source

FlowFromDataframe(String, Tuple<Int32, Int32>, String, String[], String, Int32, Boolean, Nullable<Int32>, String, String, String, Boolean, String, String)

Flows from dataframe.

Declaration
public KerasIterator FlowFromDataframe(string directory, Tuple<int, int> target_size = null, string color_mode = "rgb", string[] classes = null, string class_mode = "categorical", int batch_size = 32, bool shuffle = true, int? seed = default(int? ), string save_to_dir = "", string save_prefix = "", string save_format = "png", bool follow_links = false, string subset = "", string interpolation = "nearest")
Parameters
Type Name Description
System.String directory

string, path to the target directory. It should contain one subdirectory per class. Any PNG, JPG, BMP, PPM or TIF images inside each of the subdirectories directory tree will be included in the generator. See this script for more details.

System.Tuple<System.Int32, System.Int32> target_size

Tuple of integers (height, width), default: (256, 256). The dimensions to which all images found will be resized.

System.String color_mode

One of "grayscale", "rgb", "rgba". Default: "rgb". Whether the images will be converted to have 1, 3, or 4 channels.

System.String[] classes

Optional list of class subdirectories (e.g. ['dogs', 'cats']). Default: None. If not provided, the list of classes will be automatically inferred from the subdirectory names/structure under directory, where each subdirectory will be treated as a different class (and the order of the classes, which will map to the label indices, will be alphanumeric). The dictionary containing the mapping from class names to class indices can be obtained via the attribute class_indices.

System.String class_mode

One of "categorical", "binary", "sparse", "input", or None. Default: "categorical". Determines the type of label arrays that are returned:

System.Int32 batch_size

Size of the batches of data (default: 32).

System.Boolean shuffle

Whether to shuffle the data (default: True) If set to False, sorts the data in alphanumeric order.

System.Nullable<System.Int32> seed

Optional random seed for shuffling and transformations.

System.String save_to_dir

None or str (default: None). This allows you to optionally specify a directory to which to save the augmented pictures being generated (useful for visualizing what you are doing).

System.String save_prefix

Str. Prefix to use for filenames of saved pictures (only relevant if save_to_dir is set).

System.String save_format

One of "png", "jpeg" (only relevant if save_to_dir is set). Default: "png".

System.Boolean follow_links

Whether to follow symlinks inside class subdirectories (default: False).

System.String subset

Subset of data ("training" or "validation") if validation_split is set in ImageDataGenerator.

System.String interpolation

Interpolation method used to resample the image if the target size is different from that of the loaded image. Supported methods are "nearest", "bilinear", and "bicubic". If PIL version 1.1.3 or newer is installed, "lanczos" is also supported. If PIL version 3.4.0 or newer is installed, "box" and "hamming" are also supported. By default, "nearest" is used.

Returns
Type Description
KerasIterator

A DirectoryIterator yielding tuples of (x, y) where x is a numpy array containing a batch of images with shape (batch_size, *target_size, channels) and y is a numpy array of corresponding labels.

Implements

System.IDisposable
  • Improve this Doc
  • View Source
Back to top Generated by DocFX