Packages

object ImageTransformer extends DefaultParamsReadable[ImageTransformer] with Serializable

Pipelined image processing.

Linear Supertypes
Serializable, Serializable, DefaultParamsReadable[ImageTransformer], MLReadable[ImageTransformer], AnyRef, Any
Ordering
  1. Alphabetic
  2. By Inheritance
Inherited
  1. ImageTransformer
  2. Serializable
  3. Serializable
  4. DefaultParamsReadable
  5. MLReadable
  6. AnyRef
  7. Any
  1. Hide All
  2. Show All
Visibility
  1. Public
  2. All

Value Members

  1. def convertToTensor(matrices: Array[Mat]): Array[Array[Array[Double]]]

    Convert channel matrices to tensor in the shape of (C * H * W)

  2. def decodeImage(decodeMode: String)(r: Any): Option[(String, Mat)]

    Convert Spark image representation to OpenCV format.

  3. def encodeImage(path: String, image: Mat): Row

    Convert from OpenCV format to Dataframe Row.

  4. def extractChannels(channelOrder: String, autoConvertToColor: Boolean)(image: Mat): Array[Mat]

    Extract channels from image.

  5. def load(path: String): ImageTransformer
    Definition Classes
    ImageTransformer → MLReadable
  6. def normalizeChannels(means: Option[Array[Double]], stds: Option[Array[Double]], scaleFactor: Option[Double])(channels: Array[Mat]): Array[Mat]

    Normalize each channel.

  7. def processImage(stages: Seq[ImageTransformerStage])(image: Mat): Mat

    Apply all OpenCV transformation stages to a single image.

    Apply all OpenCV transformation stages to a single image. Break on OpenCV errors.

  8. def read: MLReader[ImageTransformer]
    Definition Classes
    DefaultParamsReadable → MLReadable