public class ImageVectorizer extends Object implements Vectorizer
Constructor and Description |
---|
ImageVectorizer(File image,
int numLabels,
int label)
Baseline knowledge needed for the vectorizer
|
Modifier and Type | Method and Description |
---|---|
ImageVectorizer |
binarize()
Binarize the data based on the threshold (anything < threshold is zero)
This is used for making the image brightness agnostic.
|
ImageVectorizer |
binarize(int threshold)
Binarize the data based on the threshold (anything < threshold is zero)
This is used for making the image brightness agnostic.
|
ImageVectorizer |
normalize()
Normalize the input image by row sums
|
org.nd4j.linalg.dataset.DataSet |
vectorize()
Vectorizes the input source in to a dataset
|
public ImageVectorizer(File image, int numLabels, int label)
image
- the input image to convertnumLabels
- the number of labelslabel
- the label of this imagepublic ImageVectorizer binarize(int threshold)
public ImageVectorizer binarize()
public ImageVectorizer normalize()
public org.nd4j.linalg.dataset.DataSet vectorize()
Vectorizer
vectorize
in interface Vectorizer
Copyright © 2015. All Rights Reserved.