Javascript Code

//////////// AOI TIFF IMAGES ///////////
var guid7 = ee.Image('projects/ee-sofiaprandelli/assets/Guidonia/orto_guidonia_gauss_07')
    .reproject('EPSG:3857');
var guid8 = ee.Image('projects/ee-sofiaprandelli/assets/Guidonia/orto_guidonia_gauss_08')
    .reproject('EPSG:3857');
var guid10 = ee.Image('projects/ee-sofiaprandelli/assets/Guidonia/orto_guidonia_gauss_10')
    .reproject('EPSG:3857');
var guid13 = ee.Image('projects/ee-sofiaprandelli/assets/Guidonia/orto_guidonia_gauss_13')
    .reproject('EPSG:3857');
/*Map.addLayer(guid7.geometry(), {color: 'FF0000'}, '7');
Map.addLayer(guid8.geometry(), {color: 'FF0000'}, '8');
Map.addLayer(guid10.geometry(), {color: 'FF0000'}, '10');
Map.addLayer(guid13.geometry(), {color: 'FF0000'}, '13');*/

var ICguid = ee.ImageCollection.fromImages([guid7, guid10, guid13]);
var guid = ICguid.mosaic();
//Map.addLayer(guid, {}, 'Mosaic');
print('Guidonia ortophoto 13, resolution:', guid13.projection().nominalScale());

print('Projection Guidonia ortophoto:', guid.projection());
print('image geometry', ICguid.geometry());
print('ICguid infos', ICguid.getInfo());

var polygon = ICguid.geometry();
Map.addLayer(polygon, {color: 'red'}, 'Study area');
print('Mosaic Coordinates:', polygon);

//////////// IMPORT RAIL COLLECTION //////////
// Tratta ferroviarie unificate e filtrate (analisi su R, file "hardDisk/eco_corridors.Rmd)
var rail = ee.FeatureCollection('projects/ee-sofiaprandelli/assets/Railway/filtered-rail');
//print(rail);
// Number of features
var count = rail.size();
print("Number of italian railways features", count);

// Filter rail in the test area
var filter_rail = rail.filterBounds(polygon);
Map.addLayer(filter_rail, {color: 'blue'}, 'Railways (test area)');
var count_filter_rail =  filter_rail.size();
print("Number of filtered railway (test area)", count_filter_rail);

var filter_rail_lenght = filter_rail.geometry().length();
print('filter rail lenght (m)', filter_rail_lenght);

// CREATE BUFFER around rail in test area
var railbuff10 = filter_rail.map(function(feature) {
  return feature.buffer(20);
});
print(railbuff10.first(), '10 m rail buffer');
print(railbuff10);
//Map.addLayer(railbuff10, {color:'yellow'}, '10 m rail buffer');

//Dissolve FC
var railbuff10_diss = railbuff10.union().first();
//print(railbuff10_diss);

// Total area in the buffer
var railbuff10_diss_area = railbuff10_diss.geometry().area();
//var railbuff10_diss_AreaSqKm = ee.Number(railbuff10_diss_area).divide(1e6).round()
print('Total area in the buffer (m2)', railbuff10_diss_area);

//DOWNSCALING TIFF IMAGE
print(guid.projection());

var guid_reduced = guid
    .reproject({
      crs: "EPSG:3857",
      scale: 1.0
    });
    
print('Guidonia reduced', guid_reduced);
print('Guidonia reduced resolution:', guid_reduced.projection().nominalScale());

// Filter raster Image by buffer zone, clip the raster image using the buffer
var guid_filter = guid_reduced.clip(railbuff10);
Map.addLayer(guid_filter, {}, 'Ortophoto Guidonia filtered');

/////////// CREATE GROUND CONTROL POINTS using GUI tool //////////

// Uncomment next lines to create standardized layers, but add property "class" to every layer
// you can also import layer with points already created from other scripts

//var rails = /* color: #a2adb1 */ee.FeatureCollection([]),
//    trees = /* color: #2b932c */ee.FeatureCollection([]),
//    grass = /* color: #6ae058 */ee.FeatureCollection([]),
//    bare = /* color: #e0df3e */ee.FeatureCollection([]),
//    shrubs = /* color: #92cc61 */ee.FeatureCollection([]);

// Set import as FeatureCollection
// Add property 'class' and assign integer value starting from 0
// Merge GCPs
var gcps = rails.merge(trees).merge(grass).merge(bare).merge(shrubs);//.merge(built);
print('GCPS', gcps);

// To assess classification accuracy, we use only 70% of the data to train our classifier. The remaining 30% is used for validation
// Add column of random numbers to the ground control points
var gcps = gcps.randomColumn();
var trainingGcps = gcps.filter(ee.Filter.lt('random', 0.7));
var validationGcps = gcps.filter(ee.Filter.gte('random', 0.7));
// Check sizes: attenzione a non sovrascrivere elementi, 
print('Size of training Gcps', trainingGcps.size());
print('Size of validation Gcps', validationGcps.size());

// Overlay the point on the image to get training data
var training = guid_filter.sampleRegions({
  collection: trainingGcps, 
  properties: ['class']/*, 
  scale: 10*/
});
print('Training data', training);
// Get band names from the image
var bands = guid_filter.bandNames();

// Train a RF classifier
var classifier = ee.Classifier.smileRandomForest(70).train({
  features: training, 
  classProperty: 'class', 
  inputProperties: bands
});

// CLASSIFY the image
var classified = guid_filter.classify(classifier);
print('Classified image', classified);
var visParamsClass = {min: 0,
  max: 4, 
  palette:['#a2adb1', '#2b932c', '#6ae058', '#e0df3e', '#92cc61']
};
//Map.addLayer(classified, visParamsClass, 'Land cover classification (not smoothed)');

// Smoothen classified map (area approximation)
var classified_smooth = classified.focalMode(1.0,'square');
//Map.addLayer(classified_smooth, visParamsClass, 'Land cover classification');

// ACCURACY ASSESSMENT: for relaibility, at least 80%
// Use classification map to assess accuracy using the validation fraction of the overall training set
var test = classified_smooth.sampleRegions({ 
  collection: validationGcps,
  properties: ['class']/*,
  scale: 10,*/
});
print('Test', test);

// CONFUSION MATRIX (performance measurement for ML classification)
var testConfusionMatrix = test.errorMatrix('class', 'classification');
// .errorMatrix() create a simple error matrix from the validation data
// The function uses the already known 'class' property and looks at the 'classification' property which represents how the classifier assigned the pixel
print('Confusion Matrix', testConfusionMatrix);
print('Test Accuracy', testConfusionMatrix.accuracy());

// POST-PROCESSING CLASSIFICATION RESULTS
// Removing noise caused by misclassified pixels: replace isolated pixels with surrounding value

// Count patch sizes
var patchsize = classified.connectedPixelCount(40, false);
// Run a majority filter
var filtered = classified.focal_mode({
    radius: 10,
    kernelType: 'square',
    units: 'meters',
}); 
// UpdateD image with majority filter where patch size is small
var connectedClassified =  classified.where(patchsize.lt(40),filtered);
Map.addLayer(connectedClassified, {min: 0, max: 4, palette: ['#a2adb1', '#2b932c', '#6ae058', '#e0df3e', '#92cc61']}, 
  'LC calssification (processed using Connected Pixels)');

///////// Feature Importance /////////

// Run .explain() to see what the classifer looks like
print('Explain classifier:', classifier.explain());

// Calculate variable importance
var importance = ee.Dictionary(classifier.explain().get('importance'));

// Calculate relative importance
var sum = importance.values().reduce(ee.Reducer.sum());

var relativeImportance = importance.map(function(key, val) {
   return (ee.Number(val).multiply(100)).divide(sum)
  });
print('Relative importance:', relativeImportance);

// Create a FeatureCollection to chart it
var importanceFc = ee.FeatureCollection([
  ee.Feature(null, relativeImportance)
]);

var chart = ui.Chart.feature.byProperty({
  features: importanceFc
}).setOptions({
      title: 'Feature Importance',
      vAxis: {title: 'Importance'},
      hAxis: {title: 'Feature'}
  });
print(chart);

//////// Calculate area by class /////////

// Create a 2 band image with the area image and the classified image
var areaImage = ee.Image.pixelArea().addBands(connectedClassified);

// Calculate areas
var areas = areaImage.reduceRegion({
      reducer: ee.Reducer.sum().group({
      groupField: 1,
      groupName: 'classification',
    }),
    geometry: railbuff10.geometry(),
    scale: 10,
    maxPixels: 1e30
    }); 
print('Total areas in buffer', areas);

var classAreas = ee.List(areas.get('groups'));

// Process results to extract the areas and create a FeatureCollection

// define a dictionary (list) with class names
var classNames = ee.Dictionary({
  '0': 'rails',
  '1': 'trees',
  '2': 'grass',
  '3': 'bare',
  '4': 'shrubs'
});

var classAreas = classAreas.map(function(item) {
  var areaDict = ee.Dictionary(item);
  var classNumber = ee.Number(areaDict.get('classification')).format();
  var className = classNames.get(classNumber);
  var area = ee.Number(
    areaDict.get('sum'));
  return ee.Feature(null, {'class': classNumber, 'class_name': className, 'area': area});
});

var classAreaFc = ee.FeatureCollection(classAreas);

// If area is large, first export the FC as an Asset and import it once the export is finished
// Bar Chart
var areaChart = ui.Chart.feature.byProperty({
  features: classAreaFc,
  xProperties: ['area'],
  seriesProperty: 'class_name',
}).setChartType('ColumnChart')
  .setOptions({
    hAxis: {title: 'Classes'},
    vAxis: {title: 'Area m^2'},
    title: 'Area by class',
    series: {
      0: { color: '#a2adb1' },
      1: { color: '#2b932c' },
      2: { color: '#6ae058' },
      3: { color: '#e0df3e' },
      4: { color: '#92cc61' }
    }
  });
print(areaChart); 

// Pie Chart
var palette = ['#a2adb1', '#2b932c', '#6ae058', '#e0df3e', '#92cc61'];

var areaChart = ui.Chart.feature.byFeature({
  features: classAreaFc,
  xProperty: 'class_name',
  yProperties: ['area']
}).setChartType('PieChart')
  .setOptions({
    hAxis: {title: 'Classes'},
    vAxis: {title: 'Area (sq m)'},
    title: 'Percentage of area by class',
    colors: palette
  });
print(areaChart);

/////////////// COPERNICUS CORINE LAND COVER LEVEL III DATA ANALYSIS ///////////////
// 44 classes

var corineIII = ee.Image('COPERNICUS/CORINE/V20/100m/2012');
var corineIIILC = corineIII.select('landcover');

// FILTER CorineLC by buffer zone in test area
var corineIII_in_buffer = corineIIILC.clip(railbuff10);
Map.addLayer(corineIII_in_buffer, {}, 'Corine LC (in buffer of 30 m)');
print('Corine LC in buffer', corineIII_in_buffer);

// AREA CALCULATION BY CLASSES
var areaImageCorineIII = ee.Image.pixelArea().addBands(corineIII_in_buffer);
print('Info Corine LC in buffer', areaImageCorineIII);

var areaCorineIII = areaImageCorineIII
    .reduceRegion({
      reducer: ee.Reducer.sum().group({ //divide(1000000). dopo sum(). per calcolo in km2
      groupField: 1,
      groupName: 'Map',
    }),
    geometry: railbuff10_diss.geometry(),
    scale: 10,
    maxPixels: 1e10
    });//.get('groups');

print('Total areas in buffer (Corine LC)', areaCorineIII);

// Process results to extract the areas and create FC
// Define a dictionary with class names
var classAreasCorineIII = ee.List(areaCorineIII.get('groups'));

var classNamesCorineIII = ee.Dictionary({
  '111':	'Continuous urban fabric',
  '112':	'Discontinuous urban fabric',
  '121':	'Industrial or commercial units',
  '122':  'Road and rail networks and associated land',
  '123':  'Port areas',
  '124':  'Airports',
  '131':  'Mineral extraction sites',
  '132':  'Dump sites',
  '133':  'mine, dump, and construction sites',
  '141':  'Green urban areas',
  '142':	'Sport and leisure facilities',
  '211':	'Non-irrigated arable land',
  '212':  'Port areas',
  '213':  'Rice fields',
  '221':  'Vineyards',
  '222':  'Fruit trees and berry plantations',
  '223':  'Olive groves',
  '231':  'Pastures',
  '241':  'Heterogeneous agricultural areas: Annual crops associated with permanent crops',
  '242':	'Complex cultivation patterns',
  '243':  'Land principally occupied by agriculture, with significant areas of natural vegetation',
  '244':  'Agro-forestry areas',
  '311':	'Forests, Broad-leaved forest',
  '312':  'Forests, Coniferous forest',
  '313':	'Forests, Mixed forest',
  '321':	'Scrub and/or herbaceous vegetation: Natural grasslands',
  '322':	'Scrub and/or herbaceous vegetation: Moors and heathland',
  '323':	'Scrub and/or herbaceous vegetation: Sclerophyllous vegetation',
  '324':	'Scrub and/or herbaceous vegetation: Transitional woodland-shrub',
  '331':  'Open spaces with little or no vegetation > Beaches, dunes, sands',
  '332':	'Open spaces with little or no vegetation > Bare rocks',
  '333':  'Open spaces with little or no vegetation > Sparsely vegetated areas',
  '334':	'Open spaces with little or no vegetation > Burnt areas',
  '335':  'Open spaces with little or no vegetation > Glaciers and perpetual snow',
  '411':  'Inland wetlands: Inland marshes',
  '412':  'Inland wetlands: Peat bogs',
  '421':	'Maritime wetlands: Salt marshes',
  '422':	'Maritime wetlands: Salines',
  '423':	'Maritime wetlands: Intertidal flats',
  '511':	'Water courses',
  '512':  'Inland waters',
  '521':	'Coastal lagoons',
  '522':	'Estuaries',
  '523':	'Sea and ocean'
});

var classAreasCorineIII = classAreasCorineIII.map(function(item) {
  var areaDictCorineIII = ee.Dictionary(item);
  var classNumberCorineIII = ee.Number(areaDictCorineIII.get('Map')).format();
  var classNameCorineIII = classNamesCorineIII.get(classNumberCorineIII);
  var areaCorineIII = ee.Number(areaDictCorineIII.get('sum'));
  return ee.Feature(null, {'class': classNumberCorineIII, 'class_name': classNameCorineIII, 'area': areaCorineIII});
});

var classAreaFcCorineIII = ee.FeatureCollection(classAreasCorineIII);

// CHART the resulting FC
// If the area is large: first Export FC as an Asset and import it once the export is finished

// Bar Chart
var areaChartCorineIII = ui.Chart.feature.byProperty({
  features: classAreaFcCorineIII,
  xProperties: ['area'],
  seriesProperty: 'class_name',
}).setChartType('ColumnChart')
  .setOptions({
    hAxis: {title: 'Classes'},
    vAxis: {title: 'Area (sq m)'},
    title: 'Area by Corine III Land Cover classes',
    series: {
        111: { color: '#e6004d'},
        112: { color: '#ff0000'},
        121: { color: '#cc4df2'},	
        122: { color: '#cc0000'},	
        123: { color: '#e6cccc'},	
        124: { color: '#e6cce6'},	
        131: { color: '#a600cc'},	
        132: { color: '#a64dcc'},	
        133: { color: '#ff4dff'},	
        141: { color: '#ffa6ff'},	
        142: { color: '#ffe6ff'},	
        211: { color: '#ffffa8'},	
        212: { color: '#ffff00'},	
        213: { color: '#e6e600'},	
        221: { color: '#e68000'},	
        222: { color: '#f2a64d'},	
        223: { color: '#e6a600'},	
        231: { color: '#e6e64d'},	
        241: { color: '#ffe6a6'},	
        242: { color: '#ffe64d'},	
        243: { color: '#e6cc4d'},	
        244: { color: '#f2cca6'},	
        311: { color: '#80ff00'},
        312: { color: '#00a600'},	
        313: { color: '#4dff00'},	
        321: { color: '#ccf24d'},	
        322: { color: '#a6ff80'},	
        323: { color: '#a6e64d'},	
        324: { color: '#a6f200'},	
        331: { color: '#e6e6e6'},
        332: { color: '#cccccc'},	
        333: { color: '#ccffcc'},	
        334: { color: '#000000'},	
        335: { color: '#a6e6cc'},	
        411: { color: '#a6a6ff'},	
        412: { color: '#4d4dff'},	
        421: { color: '#ccccff'},	
        422: { color: '#e6e6ff'},	
        423: { color: '#a6a6e6'},	
        511: { color: '#00ccf2'},	
        512: { color: '#80f2e6'},	
        521: { color: '#00ffa6'},
        522: { color: '#a6ffe6'},	
        523: { color: '#e6f2ff'},	
    }
  });
print(areaChartCorineIII); 

// Pie Chart
var paletteCorineIII = ['#e6004d', '#ff0000', '#cc4df2', '#cc0000', '#e6cccc', '#e6cce6',
'#a600cc', '#a64dcc', '#ff4dff', '#ffa6ff', '#ffe6ff', '#ffffa8', '#ffff00', '#e6e600', '#e68000', 
'#f2a64d', '#e6a600', '#e6e64d', '#ffe6a6', '#ffe64d', '#e6cc4d', '#f2cca6', '#80ff00', '#00a600',
'#4dff00', '#ccf24d', '#a6ff80','#a6e64d', '#a6f200', '#e6e6e6', '#cccccc', '#ccffcc', '#000000', 
'#a6e6cc', '#a6a6ff', '#4d4dff', '#ccccff', '#e6e6ff', '#a6a6e6', '#00ccf2', '#80f2e6', '#00ffa6',
'#a6ffe6', '#e6f2ff'
];

var areaChartCorineIII = ui.Chart.feature.byFeature({
  features: classAreaFcCorineIII,
  xProperty: 'class_name',
  yProperties: ['area']
}).setChartType('PieChart')
  .setOptions({
    hAxis: {title: 'Classes'},
    vAxis: {title: 'Area (sq m)'},
    title: 'Percentage of Area by Corine III Land Cover classes',
    series: {
        111: { color: '#e6004d'},
        112: { color: '#ff0000'},
        121: { color: '#cc4df2'},	
        122: { color: '#cc0000'},	
        123: { color: '#e6cccc'},	
        124: { color: '#e6cce6'},	
        131: { color: '#a600cc'},	
        132: { color: '#a64dcc'},	
        133: { color: '#ff4dff'},	
        141: { color: '#ffa6ff'},	
        142: { color: '#ffe6ff'},	
        211: { color: '#ffffa8'},	
        212: { color: '#ffff00'},	
        213: { color: '#e6e600'},	
        221: { color: '#e68000'},	
        222: { color: '#f2a64d'},	
        223: { color: '#e6a600'},	
        231: { color: '#e6e64d'},	
        241: { color: '#ffe6a6'},	
        242: { color: '#ffe64d'},	
        243: { color: '#e6cc4d'},	
        244: { color: '#f2cca6'},	
        311: { color: '#80ff00'},
        312: { color: '#00a600'},	
        313: { color: '#4dff00'},	
        321: { color: '#ccf24d'},	
        322: { color: '#a6ff80'},	
        323: { color: '#a6e64d'},	
        324: { color: '#a6f200'},	
        331: { color: '#e6e6e6'},
        332: { color: '#cccccc'},	
        333: { color: '#ccffcc'},	
        334: { color: '#000000'},	
        335: { color: '#a6e6cc'},	
        411: { color: '#a6a6ff'},	
        412: { color: '#4d4dff'},	
        421: { color: '#ccccff'},	
        422: { color: '#e6e6ff'},	
        423: { color: '#a6a6e6'},	
        511: { color: '#00ccf2'},	
        512: { color: '#80f2e6'},	
        521: { color: '#00ffa6'},
        522: { color: '#a6ffe6'},	
        523: { color: '#e6f2ff'},	
    }
  });
print(areaChartCorineIII);