Main Content

randomPatchExtractionDatastore

Datastore for extracting random 2-D or 3-D random patches from images or pixel label images

expand all in page

Description

一个randomPatchExtractionDatastoreextracts corresponding randomly-positioned patches from two image-based datastores. For example, the input datastores can be two image datastores that contain the network inputs and desired network responses for training image-to-image regression networks, or ground truth images and pixel label data for training semantic segmentation networks.

This object requires that you have Deep Learning Toolbox™.

Note

When you use arandomPatchExtractionDatastoreas a source of training data, the datastore extracts multiple random patches from each image for each epoch, so that each epoch uses a slightly different data set. The actual number of training patches at each epoch is the number of training images multiplied byPatchesPerImage。图像补丁不存储在内存中。

Creation

Syntax

patchds = randomPatchExtractionDatastore(ds1,ds2,PatchSize)
patchds = randomPatchExtractionDatastore(ds1,ds2,PatchSize,Name,Value)

Description

example

patchds = randomPatchExtractionDatastore(ds1,ds2,PatchSize)creates a datastore that extracts randomly-positioned patches of sizePatchSizefrom input data in datastoreds1and response data in datastoreds2。

patchds = randomPatchExtractionDatastore(ds1,ds2,PatchSize,Name,Value)uses name-value pairs to set thePatchesPerImage,DataAugmentation, andDispatchInBackgroundproperties. You can specify multiple name-value pairs. Enclose each property name in quotes.

For example,randomPatchExtractionDatastore(imds1,imds2,50,'PatchesPerImage',40)creates a datastore that randomly generates 40 patches of size 50-by-50 pixels from each image in image datastoresimds1andimds2。

Input Arguments

expand all

Input data containing training input to the network, specified as anImageDatastore,PixelLabelDatastore(计算机矢量绘图软件n Toolbox), orTransformedDatastore。

Specifying aPixelLabelDatastorerequires Computer Vision Toolbox™.

Note

ImageDatastoreallows batch-reading of JPG or PNG image files using prefetching. If you use a custom function for reading the images, then prefetching does not happen.

Response data representing the desired network responses, specified as anImageDatastore,PixelLabelDatastore(计算机矢量绘图软件n Toolbox), orTransformedDatastore。If you specify aTransformedDatastore, then the underlying datastore must be anImageDatastoreor aPixelLabelDatastore。

Specifying aPixelLabelDatastorerequires Computer Vision Toolbox.

Note

ImageDatastoreallows batch-reading of JPG or PNG image files using prefetching. If you use a custom function for reading the images, then prefetching does not happen.

Properties

expand all

This property is read-only.

Patch size, specified as one of the following.

  • 一个2-element vector of positive integers for 2-D patches.PatchSizehas the form [rc] whererspecifies the number of rows andcspecifies the number of columns in the patch.

  • 一个3-element vector of positive integers for 3-D patches.PatchSizehas the form [rcp] whererspecifies the number of rows,cspecifies the number of columns, andpspecifies the number of planes in the patch.

Number of random patches per image, specified as a positive integer.

Preprocessing applied to input images, specified as animageDataAugmenter(Deep Learning Toolbox)object or'none'。WhenDataAugmentationis'none', no preprocessing is applied to input images.

一个ugment data with random transformations, such as resizing, rotation, and reflection, to help prevent the network from overfitting and memorizing the exact details of the training data. TherandomPatchExtractionDatastoreapplies the same random transformation to both patches in each pair. The datastore augments data in real-time while training.

TheDataAugmentationproperty is not supported for 3-D data. To preprocess 3-D data, use thetransformfunction.

Dispatch observations in the background during training, prediction, or classification, specified asfalseortrue。To use background dispatching, you must have Parallel Computing Toolbox™.

Number of observations that are returned in each batch. You can change the value ofMiniBatchSizeonly after you create the datastore. For training, prediction, and classification, theMiniBatchSizeproperty is set to the mini-batch size defined intrainingOptions(Deep Learning Toolbox)。

This property is read-only.

Total number of observations in therandomPatchExtractionDatastore。The number of observations is the length of one training epoch.

Object Functions

combine Combine data from multiple datastores
hasdata Determine if data is available to read
numpartitions Number of datastore partitions
partition Partition a datastore
partitionByIndex PartitionrandomPatchExtractionDatastoreaccording to indices
preview Preview subset of data in datastore
read Read data fromrandomPatchExtractionDatastore
readall Read all data in datastore
readByIndex Read data specified by index fromrandomPatchExtractionDatastore
reset Reset datastore to initial state
shuffle Shuffle data in datastore
transform Transform datastore
isPartitionable Determine whether datastore is partitionable
isShuffleable Determine whether datastore is shuffleable

Examples

collapse all

This example uses:

Open Live Script

Create an image datastore containing training images. The datastore in this example contains JPEG color images.

imageDir = fullfile(toolboxdir('images'),'imdata'); imds1 = imageDatastore(imageDir,'FileExtensions','.jpg');

Create a second datastore that transforms the images inimds1by applying a Gaussian blur.

imds2 = transform(imds1,@(x)imgaussfilt(x,2));

Create animageDataAugmenterthat rotates images by random angles in the range [0, 90] degrees and randomly reflects image data horizontally.

augmenter = imageDataAugmenter('RandRotation',[0 90],'RandXReflection',true)
augmenter = imageDataAugmenter with properties: FillValue: 0 RandXReflection: 1 RandYReflection: 0 RandRotation: [0 90] RandScale: [1 1] RandXScale: [1 1] RandYScale: [1 1] RandXShear: [0 0] RandYShear: [0 0] RandXTranslation: [0 0] RandYTranslation: [0 0]

Create arandomPatchExtractionDatastoreobject that extracts random patches of size 100-by-100 from the unprocessed training images and corresponding smoothed response images. Specify the augmentation options by setting theDataAugmentationproperty.

patchds = randomPatchExtractionDatastore(imds1,imds2,[100 100],。..'DataAugmentation',augmenter)
patchds = randomPatchExtractionDatastore with properties: PatchesPerImage: 128 PatchSize: [100 100] DataAugmentation: [1×1 imageDataAugmenter] MiniBatchSize: 128 NumObservations: [] DispatchInBackground: 0

Preview a set of augmented image patches and the corresponding smoothed image patches.

minibatch = preview(patchds); inputs = minibatch.InputImage; responses = minibatch.ResponseImage; test = cat(2,inputs,responses); montage(test','Size',[8 2]) title(的输入(左)和反应(右))

This example uses:

Open Live Script

Create an image datastore containing training images.

dataDir = fullfile(toolboxdir('vision'),'visiondata','triangleImages'); imageDir = fullfile(dataDir,'trainingImages'); imds = imageDatastore(imageDir);

Define class names and their associated label IDs. Then, create a pixel label datastore containing the ground truth pixel labels for the training images.

classNames = [“三角形”,"background"]; labelIDs = [255 0]; labelDir = fullfile(dataDir,'trainingLabels'); pxds = pixelLabelDatastore(labelDir,classNames,labelIDs);

Create a random patch extraction datastore to extract random patches of size 32-by-32 pixels from the images and corresponding pixel labels. Set the optionalPatchesPerImageproperty to extract 512 random patches from each image and pixel label pair.

patchds = randomPatchExtractionDatastore(imds,pxds,32,。..'PatchesPerImage',512);

Create a network for semantic segmentation.

layers = [ imageInputLayer([32 32 1]) convolution2dLayer(3,64,“填充”,1) reluLayer() maxPooling2dLayer(2,'Stride',2) convolution2dLayer(3,64,“填充”,1) reluLayer() transposedConv2dLayer(4,64,'Stride',2,'Cropping',1) convolution2dLayer(1,2) softmaxLayer() pixelClassificationLayer() ]
layers = 10x1 Layer array with layers: 1 '' Image Input 32x32x1 images with 'zerocenter' normalization 2 '' Convolution 64 3x3 convolutions with stride [1 1] and padding [1 1 1 1] 3 '' ReLU ReLU 4 '' Max Pooling 2x2 max pooling with stride [2 2] and padding [0 0 0 0] 5 '' Convolution 64 3x3 convolutions with stride [1 1] and padding [1 1 1 1] 6 '' ReLU ReLU 7 '' Transposed Convolution 64 4x4 transposed convolutions with stride [2 2] and output cropping [1 1] 8 '' Convolution 2 1x1 convolutions with stride [1 1] and padding [0 0 0 0] 9 '' Softmax softmax 10 '' Pixel Classification Layer Cross-entropy loss

Set up training options. To reduce training time, setMaxEpochsto 5.

options = trainingOptions('sgdm',。..'InitialLearnRate',1e-3,。..'MaxEpochs',5,。..'Verbose',false);

Train the network.

net = trainNetwork(patchds,layers,options);

Tips

  • TherandomPatchExtractionDatastoreexpects that the output from thereadoperation on the input datastores return arrays of the same size.

  • If the input datastore is anImageDatastore, then the values in itsLabelsproperty are ignored by therandomPatchExtractionDatastore。

  • To visualize 2-D data in arandomPatchExtractionDatastore, you can use thepreviewfunction, which returns a subset of data in a table. Visualize all of the patches in the same figure by using themontagefunction. For example, this code displays a preview of image patches from arandomPatchExtractionDatastorecalledpatchds。

    minibatch = preview(patchds); montage(minibatch.InputImage)

Version History

Introduced in R2018b

See Also

(Deep Learning Toolbox)|(计算机矢量绘图软件n Toolbox)||(计算机矢量绘图软件n Toolbox)|(Deep Learning Toolbox)|(Deep Learning Toolbox)|

Topics