Uncomment the following line to install geemap if needed.

In [1]:

# !pip install geemap

# !pip install geemap

Machine Learning with Earth Engine - Unsupervised Classification¶

Unsupervised classification algorithms available in Earth Engine¶

Source: https://developers.google.com/earth-engine/clustering

The ee.Clusterer package handles unsupervised classification (or clustering) in Earth Engine. These algorithms are currently based on the algorithms with the same name in Weka. More details about each Clusterer are available in the reference docs in the Code Editor.

Clusterers are used in the same manner as classifiers in Earth Engine. The general workflow for clustering is:

1. Assemble features with numeric properties in which to find clusters.
2. Instantiate a clusterer. Set its parameters if necessary.
3. Train the clusterer using the training data.
4. Apply the clusterer to an image or feature collection.
5. Label the clusters.

The training data is a FeatureCollection with properties that will be input to the clusterer. Unlike classifiers, there is no input class value for an Clusterer. Like classifiers, the data for the train and apply steps are expected to have the same number of values. When a trained clusterer is applied to an image or table, it assigns an integer cluster ID to each pixel or feature.

Here is a simple example of building and using an ee.Clusterer:

Step-by-step tutorial¶

Import libraries¶

In [2]:

import ee
import geemap

import ee import geemap

Create an interactive map¶

In [3]:

Map = geemap.Map()
Map

Map = geemap.Map() Map

Out[3]:

Make this Notebook Trusted to load map: File -> Trust Notebook

Add data to the map¶

In [4]:

# point = ee.Geometry.Point([-122.4439, 37.7538])
point = ee.Geometry.Point([-87.7719, 41.8799])

image = (
    ee.ImageCollection('LANDSAT/LC08/C01/T1_SR')
    .filterBounds(point)
    .filterDate('2019-01-01', '2019-12-31')
    .sort('CLOUD_COVER')
    .first()
    .select('B[1-7]')
)

vis_params = {'min': 0, 'max': 3000, 'bands': ['B5', 'B4', 'B3']}

Map.centerObject(point, 8)
Map.addLayer(image, vis_params, "Landsat-8")

# point = ee.Geometry.Point([-122.4439, 37.7538]) point = ee.Geometry.Point([-87.7719, 41.8799]) image = ( ee.ImageCollection('LANDSAT/LC08/C01/T1_SR') .filterBounds(point) .filterDate('2019-01-01', '2019-12-31') .sort('CLOUD_COVER') .first() .select('B[1-7]') ) vis_params = {'min': 0, 'max': 3000, 'bands': ['B5', 'B4', 'B3']} Map.centerObject(point, 8) Map.addLayer(image, vis_params, "Landsat-8")

Check image properties¶

In [5]:

props = geemap.image_props(image)
props.getInfo()

props = geemap.image_props(image) props.getInfo()

Out[5]:

{'CLOUD_COVER': 0.03,
 'CLOUD_COVER_LAND': 0.04,
 'EARTH_SUN_DISTANCE': 1.016591,
 'ESPA_VERSION': '2_23_0_1b',
 'GEOMETRIC_RMSE_MODEL': 6.348,
 'GEOMETRIC_RMSE_MODEL_X': 4.429,
 'GEOMETRIC_RMSE_MODEL_Y': 4.547,
 'IMAGE_DATE': '2019-07-12',
 'IMAGE_QUALITY_OLI': 9,
 'IMAGE_QUALITY_TIRS': 9,
 'LANDSAT_ID': 'LC08_L1TP_022031_20190712_20190719_01_T1',
 'LEVEL1_PRODUCTION_DATE': 1563565308000,
 'NOMINAL_SCALE': 30,
 'PIXEL_QA_VERSION': 'generate_pixel_qa_1.6.0',
 'SATELLITE': 'LANDSAT_8',
 'SENSING_TIME': '2019-07-12T16:28:51.3794760Z',
 'SOLAR_AZIMUTH_ANGLE': 131.949371,
 'SOLAR_ZENITH_ANGLE': 26.494972,
 'SR_APP_VERSION': 'LaSRC_1.3.0',
 'WRS_PATH': 22,
 'WRS_ROW': 31,
 'system:asset_size': '553.046839 MB',
 'system:band_names': ['B1', 'B2', 'B3', 'B4', 'B5', 'B6', 'B7'],
 'system:id': 'LANDSAT/LC08/C01/T1_SR/LC08_022031_20190712',
 'system:index': 'LC08_022031_20190712',
 'system:time_end': '2019-07-12 16:28:51',
 'system:time_start': '2019-07-12 16:28:51',
 'system:version': 1564390084630490.0}

In [6]:

props.get('IMAGE_DATE').getInfo()

props.get('IMAGE_DATE').getInfo()

Out[6]:

'2019-07-12'

In [7]:

props.get('CLOUD_COVER').getInfo()

props.get('CLOUD_COVER').getInfo()

Out[7]:

0.03

Make training dataset¶

There are several ways you can create a region for generating the training dataset.

• Draw a shape (e.g., rectangle) on the map and the use region = Map.user_roi
• Define a geometry, such as region = ee.Geometry.Rectangle([-122.6003, 37.4831, -121.8036, 37.8288])
• Create a buffer zone around a point, such as region = ee.Geometry.Point([-122.4439, 37.7538]).buffer(10000)
• If you don't define a region, it will use the image footprint by default

In [8]:

# region = Map.user_roi
# region = ee.Geometry.Rectangle([-122.6003, 37.4831, -121.8036, 37.8288])
# region = ee.Geometry.Point([-122.4439, 37.7538]).buffer(10000)

# region = Map.user_roi # region = ee.Geometry.Rectangle([-122.6003, 37.4831, -121.8036, 37.8288]) # region = ee.Geometry.Point([-122.4439, 37.7538]).buffer(10000)

In [9]:

# Make the training dataset.
training = image.sample(
    **{
        #     'region': region,
        'scale': 30,
        'numPixels': 5000,
        'seed': 0,
        'geometries': True,  # Set this to False to ignore geometries
    }
)

Map.addLayer(training, {}, 'training', False)
Map

# Make the training dataset. training = image.sample( **{ # 'region': region, 'scale': 30, 'numPixels': 5000, 'seed': 0, 'geometries': True, # Set this to False to ignore geometries } ) Map.addLayer(training, {}, 'training', False) Map

Out[9]:

Make this Notebook Trusted to load map: File -> Trust Notebook

Train the clusterer¶

In [10]:

# Instantiate the clusterer and train it.
n_clusters = 5
clusterer = ee.Clusterer.wekaKMeans(n_clusters).train(training)

# Instantiate the clusterer and train it. n_clusters = 5 clusterer = ee.Clusterer.wekaKMeans(n_clusters).train(training)

Classify the image¶

In [11]:

# Cluster the input using the trained clusterer.
result = image.cluster(clusterer)

# # Display the clusters with random colors.
Map.addLayer(result.randomVisualizer(), {}, 'clusters')
Map

# Cluster the input using the trained clusterer. result = image.cluster(clusterer) # # Display the clusters with random colors. Map.addLayer(result.randomVisualizer(), {}, 'clusters') Map

Out[11]:

Make this Notebook Trusted to load map: File -> Trust Notebook

Label the clusters¶

In [12]:

legend_keys = ['One', 'Two', 'Three', 'Four', 'ect']
legend_colors = ['#8DD3C7', '#FFFFB3', '#BEBADA', '#FB8072', '#80B1D3']

# Reclassify the map
result = result.remap([0, 1, 2, 3, 4], [1, 2, 3, 4, 5])

Map.addLayer(
    result, {'min': 1, 'max': 5, 'palette': legend_colors}, 'Labelled clusters'
)
Map.add_legend(
    legend_keys=legend_keys, legend_colors=legend_colors, position='bottomright'
)
Map

legend_keys = ['One', 'Two', 'Three', 'Four', 'ect'] legend_colors = ['#8DD3C7', '#FFFFB3', '#BEBADA', '#FB8072', '#80B1D3'] # Reclassify the map result = result.remap([0, 1, 2, 3, 4], [1, 2, 3, 4, 5]) Map.addLayer( result, {'min': 1, 'max': 5, 'palette': legend_colors}, 'Labelled clusters' ) Map.add_legend( legend_keys=legend_keys, legend_colors=legend_colors, position='bottomright' ) Map

Out[12]:

Make this Notebook Trusted to load map: File -> Trust Notebook

Visualize the result¶

In [13]:

print('Change layer opacity:')
cluster_layer = Map.layers[-1]
cluster_layer.interact(opacity=(0, 1, 0.1))

print('Change layer opacity:') cluster_layer = Map.layers[-1] cluster_layer.interact(opacity=(0, 1, 0.1))

Change layer opacity:

---------------------------------------------------------------------------
AttributeError                            Traceback (most recent call last)
Input In [13], in <cell line: 2>()
      1 print('Change layer opacity:')
----> 2 cluster_layer = Map.layers[-1]
      3 cluster_layer.interact(opacity=(0, 1, 0.1))

AttributeError: 'Map' object has no attribute 'layers'

Export the result¶

Export the result directly to your computer:

In [14]:

import os

out_dir = os.path.join(os.path.expanduser('~'), 'Downloads')
out_file = os.path.join(out_dir, 'cluster.tif')

import os out_dir = os.path.join(os.path.expanduser('~'), 'Downloads') out_file = os.path.join(out_dir, 'cluster.tif')

In [15]:

geemap.ee_export_image(result, filename=out_file, scale=90)

geemap.ee_export_image(result, filename=out_file, scale=90)

Generating URL ...
Downloading data from https://earthengine.googleapis.com/v1alpha/projects/earthengine-legacy/thumbnails/18979d5b09def11ea51d29891e017efd-f01024a70f3a8539e679655e4194c76e:getPixels
Please wait ...
Data downloaded to /home/runner/Downloads/cluster.tif

Export the result to Google Drive:

In [16]:

geemap.ee_export_image_to_drive(
    result, description='clusters', folder='export', scale=90
)

geemap.ee_export_image_to_drive( result, description='clusters', folder='export', scale=90 )

Exporting clusters ...

---

Last update: 2022-03-14