You can run this notebook in a live session Binder or view it on Github.

Image Processing

Welcome to the quickstart guide for dask-image.

  1. Setting up your environment

  2. Importing dask-image

  3. Getting the example data

  4. Reading in image data

    1. Reading a single image

    2. Reading multiple iamges

  5. Applying your own custom function to images

    1. Embarrassingly parallel problems

  6. Joining partial images together

  7. A segmentation analysis pipeline

    1. Filtering

    2. Segmenting

    3. Analyzing

  8. Next steps

  9. Cleaning up temporary directories and files

## Setting up your environment

Install Extra Dependencies

We first install the library scikit-image for easier access to the example image data there.

!pip install scikit-image --upgrade-strategy only-if-needed
Collecting scikit-image
  Downloading (26.3MB)

Collecting imageio>=2.0.1 (from scikit-image)
  Downloading (3.3MB)

Requirement already satisfied: matplotlib!=3.0.0,>=2.0.0 in /home/travis/miniconda/envs/test/lib/python3.6/site-packages (from scikit-image) (3.0.3)
Requirement already satisfied: pillow>=4.3.0 in /home/travis/miniconda/envs/test/lib/python3.6/site-packages (from scikit-image) (6.0.0)
Collecting PyWavelets>=0.4.0 (from scikit-image)
  Downloading (4.4MB)

Requirement already satisfied: scipy>=0.17.0 in /home/travis/miniconda/envs/test/lib/python3.6/site-packages (from scikit-image) (1.2.1)
Collecting networkx>=2.0 (from scikit-image)
  Downloading (1.7MB)

Requirement already satisfied: numpy in /home/travis/miniconda/envs/test/lib/python3.6/site-packages (from imageio>=2.0.1->scikit-image) (1.16.2)
Requirement already satisfied: cycler>=0.10 in /home/travis/miniconda/envs/test/lib/python3.6/site-packages (from matplotlib!=3.0.0,>=2.0.0->scikit-image) (0.10.0)
Requirement already satisfied: kiwisolver>=1.0.1 in /home/travis/miniconda/envs/test/lib/python3.6/site-packages (from matplotlib!=3.0.0,>=2.0.0->scikit-image) (1.0.1)
Requirement already satisfied: pyparsing!=2.0.4,!=2.1.2,!=2.1.6,>=2.0.1 in /home/travis/miniconda/envs/test/lib/python3.6/site-packages (from matplotlib!=3.0.0,>=2.0.0->scikit-image) (2.4.0)
Requirement already satisfied: python-dateutil>=2.1 in /home/travis/miniconda/envs/test/lib/python3.6/site-packages (from matplotlib!=3.0.0,>=2.0.0->scikit-image) (2.8.0)
Requirement already satisfied: decorator>=4.3.0 in /home/travis/miniconda/envs/test/lib/python3.6/site-packages (from networkx>=2.0->scikit-image) (4.4.0)
Requirement already satisfied: six in /home/travis/miniconda/envs/test/lib/python3.6/site-packages (from cycler>=0.10->matplotlib!=3.0.0,>=2.0.0->scikit-image) (1.12.0)
Requirement already satisfied: setuptools in /home/travis/miniconda/envs/test/lib/python3.6/site-packages (from kiwisolver>=1.0.1->matplotlib!=3.0.0,>=2.0.0->scikit-image) (41.0.0)
Building wheels for collected packages: networkx
  Building wheel for networkx ( ... - \ | / - done
  Stored in directory: /home/travis/.cache/pip/wheels/de/63/64/3699be2a9d0ccdb37c7f16329acf3863fd76eda58c39c737af
Successfully built networkx
Installing collected packages: imageio, PyWavelets, networkx, scikit-image
Successfully installed PyWavelets-1.0.3 imageio-2.5.0 networkx-2.3 scikit-image-0.15.0

If you are running this notebook on your own computer and not in the mybinder environment, you’ll additionally need to ensure your Python environment contains: * dask * dask-image * python-graphviz * scikit-image * matplotlib * numpy

You can refer to the full list of dependencies used for the dask-examples repository, available in the `binder/environment.yml file here <>`__ (note that the nomkl package is not available for Windows users):

## Importing dask-image

When you import dask-image, be sure to use an underscore instead of a dash between the two words.

import dask_image.imread
import dask_image.ndfilters
import dask_image.ndmeasure
import dask.array as da

We’ll also use matplotlib to display image results in this notebook.

import matplotlib.pyplot as plt
%matplotlib inline

## Getting the example data

We’ll use some example image data from the scikit-image library in this tutorial. These images are very small, but will allow us to demonstrate the functionality of dask-image.

Let’s download and save a public domain image of the astronaut Eileen Collins to a temporary directory. This image was originally downloaded from the NASA Great Images database, but we’ll access it with scikit-image’s data.astronaut() method.

!mkdir temp
import os
from skimage import data, io

output_filename = os.path.join('temp', 'astronaut.png')
io.imsave(output_filename, data.astronaut())

Really large datasets often can’t fit all of the data into a single file, so we’ll chop this image into four parts and save the image tiles to a second temporary directory. This will give you a better idea of how you can use dask-image on a real dataset.

!mkdir temp-tiles
io.imsave(os.path.join('temp-tiles', 'image-00.png'), data.astronaut()[:256, :256, :])  # top left
io.imsave(os.path.join('temp-tiles', 'image-01.png'), data.astronaut()[:256, 256:, :])  # top right
io.imsave(os.path.join('temp-tiles', 'image-10.png'), data.astronaut()[256:, :256, :])  # bottom left
io.imsave(os.path.join('temp-tiles', 'image-11.png'), data.astronaut()[256:, 256:, :])  # bottom right

Now we have some data saved, let’s practise reading in files with dask-image and processing our images.

## Reading in image data

### Reading a single image

Let’s load a public domain image of the astronaut Eileen Collins with dask-image imread(). This image was originally downloaded from the NASA Great Images database

import os
filename = os.path.join('temp', 'astronaut.png')
astronaut = dask_image.imread.imread(filename)
plt.imshow(astronaut[0, ...])  # display the first (and only) frame of the image
dask.array<from-value, shape=(1, 512, 512, 3), dtype=uint8, chunksize=(1, 512, 512, 3)>
<matplotlib.image.AxesImage at 0x7f8a934594e0>

This has created a dask array with shape=(1, 512, 512, 3). This means it contains one image frame with 512 rows, 512 columns, and 3 color channels.

Since the image is relatively small, it fits entirely within one dask-image chunk, with chunksize=(1, 512, 512, 3).

### Reading multiple images

In many cases, you may have multiple images stored on disk, for example: image_00.png, image_01.png, … image_NN.png. These can be read into a dask array as multiple image frames.

Here we have the astronaut image split into four non-overlapping tiles: * image_00.png = top left image (index 0,0) * image_01.png = top right image (index 0,1) * image_10.png = bottom left image (index 1,0) * image_11.png = bottom right image (index 1,1)

This filename pattern can be matched with regex: image-*.png

!ls temp-tiles
image-00.png  image-01.png  image-10.png  image-11.png
import os
filename_pattern = os.path.join('temp-tiles', 'image-*.png')
tiled_astronaut_images = dask_image.imread.imread(filename_pattern)
dask.array<concatenate, shape=(4, 256, 256, 3), dtype=uint8, chunksize=(1, 256, 256, 3)>

This has created a dask array with shape=(4, 256, 256, 3). This means it contains four image frames; each with 256 rows, 256 columns, and 3 color channels.

There are four chunks in this particular case. Each image frame here is a separate chunk with chunksize=(1, 256, 256, 3).

fig, ax = plt.subplots(nrows=2, ncols=2)

## Applying your own custom function to images

Next you’ll want to do some image processing, and apply a function to your images.

We’ll use a very simple example: converting an RGB image to grayscale. But you can also use this method to apply arbittrary functions to dask iamges. To convert our image to grayscale, we’ll use the equation to calcuate luminance (reference pdf)”:

Y = 0.2125 R + 0.7154 G + 0.0721 B

We’ll write the function for this equation as follows:

def grayscale(rgb):
    result = ((rgb[..., 0] * 0.2125) +
              (rgb[..., 1] * 0.7154) +
              (rgb[..., 2] * 0.0721))
    return result

Let’s apply this function to the astronaut image we read in as a single file and visualize the computation graph.

(Visualizing the computation graph isn’t necessary most of the time but it’s helpful to know what dask is doing under the hood, and it can also be very useful for debugging problems.)

single_image_result = grayscale(astronaut)
dask.array<add, shape=(1, 512, 512), dtype=float64, chunksize=(1, 512, 512)>

We also see that there are no longer three color channels in the shape of the result, and that the output image is as expected.

print("Original image dimensions: ", astronaut.shape)
print("Processed image dimensions:", single_image_result.shape)

fig, (ax0, ax1) = plt.subplots(nrows=1, ncols=2)
ax0.imshow(astronaut[0, ...])            # display the first (and only) frame of the image
ax1.imshow(single_image_result[0, ...], cmap='gray')  # display the first (and only) frame of the image

# Subplot headings
ax0.set_title('Original iamge')
ax1.set_title('Processed iamge')

# Don't display axes

# Display images
Original image dimensions:  (1, 512, 512, 3)
Processed image dimensions: (1, 512, 512)

### Embarrassingly parallel problems

The syntax is identical to apply a function to multiple images or dask chunks. This is an example of an embarrassingly parallel problem, and we see that dask automatically creates a computation graph for each chunk.

result = grayscale(tiled_astronaut_images)
dask.array<add, shape=(4, 256, 256), dtype=float64, chunksize=(1, 256, 256)>

Let’s take a look at the results.

fig, ((ax0, ax1), (ax2, ax3)) = plt.subplots(nrows=2, ncols=2)
ax0.imshow(result[0, ...], cmap='gray')
ax1.imshow(result[1, ...], cmap='gray')
ax2.imshow(result[2, ...], cmap='gray')
ax3.imshow(result[3, ...], cmap='gray')

# Subplot headings
ax0.set_title('First chunk')
ax1.set_title('Second chunk')
ax2.set_title('Thurd chunk')
ax3.set_title('Fourth chunk')

# Don't display axes

# Display images

## Joining partial images together

Ok, Things are looking pretty good! But how can we join these image chunks together?

So far, we haven’t needed any information from neighboring pixels to do our calculations. But there are lots of functions (like those in dask-image ndfilters) that do need this for accurate results. You could end up with unwanted edge effects if you don’t tell dask how your images should be joined.

Dask has several ways to join chunks together: Stack, Concatenate, and Block.

Block is very versatile, so we’ll use that in this next example. You simply pass in a list (or list of lists) to tell dask the spatial relationship between image chunks.

data = [[result[0, ...], result[1, ...]],
        [result[2, ...], result[3, ...]]]
combined_image = da.block(data)
plt.imshow(combined_image, cmap='gray')
(512, 512)
<matplotlib.image.AxesImage at 0x7f8a900f8908>

## A segmentation analysis pipeline

We’ll walk through a simple image segmentation and analysis pipeline with three steps: 1. Filtering 1. Segmenting 1. Analyzing

### Filtering

Most analysis pipelines require some degree of image preprocessing. dask-image has a number of inbuilt filters available via dask-image ndfilters

Commonly a guassian filter may used to smooth the image before segmentation. This causes some loss of sharpness in the image, but can improve segmentation quality for methods that rely on image thresholding.

smoothed_image = dask_image.ndfilters.gaussian_filter(combined_image, sigma=[1, 1])

We see a small amount of blur in the smoothed image.

fig, (ax0, ax1) = plt.subplots(nrows=1, ncols=2)
ax0.imshow(smoothed_image, cmap='gray')
ax1.imshow(smoothed_image - combined_image, cmap='gray')

# Subplot headings
ax0.set_title('Smoothed iamge')
ax1.set_title('Difference from original')

# Don't display axes

# Display images

Since the gaussian filter uses information from neighbouring pixels, the computational graph looks more complicated than the ones we looked at earlier. This is no longer embarrassingly parallel. Where possible dask keeps the computations for each of the four image chunks separate, but must combine information from different chunks near the edges.


### Segmenting

After the image preprocessing, we segment regions of interest from the data. We’ll use a simple arbitrary threshold as the cutoff, at 75% of the maximum intensity of the smoothed image.

threshold_value = 0.75 * da.max(smoothed_image).compute()
threshold_image = smoothed_image > threshold_value
plt.imshow(threshold_image, cmap='gray')
<matplotlib.image.AxesImage at 0x7f8a7afe0780>

Next, we label each region of connected pixels above the threshold value. For this we use the label function from dask-image ndmeasure. This will return both the label image, and the number of labels.

label_image, num_labels = dask_image.ndmeasure.label(threshold_image)
/home/travis/miniconda/envs/test/lib/python3.6/site-packages/dask_image/ndmeasure/ RuntimeWarning: ``input`` does not have 1 chunk in all dimensions; it will be consolidated first
  warn("``input`` does not have 1 chunk in all dimensions; it will be consolidated first", RuntimeWarning)
print("Number of labels:", int(num_labels))
plt.imshow(label_image, cmap='viridis')
Number of labels: 78
<matplotlib.image.AxesImage at 0x7f8a7af49cc0>

### Analyzing

There are a number of inbuilt functions in dask-image ndmeasure useful for quantitative analysis.

We’ll use the dask_image.ndmeasure.mean() and dask_image.ndmeasure.standard_deviation() functions, and apply them to each label region with dask_image.ndmeasure.labeled_comprehension().

index = list(range(int(num_labels)))  # Note that we're including the background label=0 here, too.
out_dtype = float  # The data type we want to use for our results.
default = None     # The value to return if an element of index does not exist in the label image.
mean_values = dask_image.ndmeasure.labeled_comprehension(combined_image, label_image, index, dask_image.ndmeasure.mean, out_dtype, default, pass_positions=False)
[ 90.87492186 194.69185981 207.76964212 215.16959146 196.92609412
 206.50032105 224.71301613 249.08085    217.0948     215.87763779
 211.0477766  205.68146    208.61031956 198.27655577 212.43803803
 201.39967857 210.291275   198.3529809  201.21037636 214.13041176
 200.9344975  201.85547778 194.46485714 202.60302231 205.20927983
 203.72510602 205.94798756 205.88514047 213.67658483 207.33728
 216.13305227 206.7058     211.6957     205.65810758 211.75921279
 197.46586082 202.35786579 199.16034783 216.27194848 200.69137594
 220.99142573 220.1454     239.55       246.5        241.53348
 236.28736    244.84466624 242.09528    203.37236667 209.34061875
 213.76621346 247.53468249 207.66165    234.37771667 224.00188182
 229.2705     232.9163     187.1873     236.16183793 223.7469
 187.813475   227.4778     244.17155    225.49225806 239.4951
 218.7795     252.488826   208.5659     242.99015132 232.4975218
 197.72356    201.98308947 230.57158889 212.82135217 203.13211
 242.21528571 241.32428889 197.91741528]

Since we’re including label 0 in our index, it’s not surprising that the first mean value is so much lower than the others - it’s the background region below our cutoff threshold for segmentation.

Let’s also calculate the standard deviation of the pixel values in our greyscale iamge.

stdev_values = dask_image.ndmeasure.labeled_comprehension(combined_image, label_image, index, dask_image.ndmeasure.standard_deviation, out_dtype, default, pass_positions=False)

Finally, let’s load our analysis results into a pandas table and then save it as a csv file.

import pandas as pd

df = pd.DataFrame()
df['label'] = index
df['mean'] = mean_values.compute()
df['standard_deviation'] = stdev_values.compute()

label mean standard_deviation
0 0 90.874922 65.452828
1 1 194.691860 2.921235
2 2 207.769642 11.411058
3 3 215.169591 9.193374
4 4 196.926094 5.215053
print('Saved example_analysis_results.csv')
Saved example_analysis_results.csv

## Next steps

I hope this guide has helped you to get started with dask-image.


You can read more about dask-image in the dask-image documentation and API reference. Documentation for dask is here.

The dask-examples repository has a number of other example notebooks:

Scaling up with dask distributed

If you want to send dask jobs to a computing cluster for distributed processing, you should take a look at dask distributed. There is also a quickstart guide available.

Saving image data with zarr

In some cases it may be necessary to save large data after image processing, zarr is a python library that you may find useful.

## Cleaning up temporary directories and files

You recall we saved some example data to the directories temp/ and temp-tiles/. To delete the contents, run the following command:

!rm -r temp
!rm -r temp-tiles