GEE Select images that have certain number of pixels

Question

I have code that calculate NDVI for Landsat 5,7 and 8. I would like to know how can I "filter" my image collection so I have only images that have a certain number of pixels. For that I thought maybe I can first count the number of pixels in the image and then use gt but I'm not sure how to put it inside the imageCollection.

To do that I think I need to take my imageCollection and change it into a list and then iterate through this list- for every image to count the numbers of pixels using something like this-

//count the number of total pixels
var countpixels = image1.reduceRegion({
  reducer: ee.Reducer.count(),
  geometry: table,
  crs: 'EPSG:4326',
  scale: 30,
  });

and then filter according to the number of pixels.

My problems are-

1. I have around 250 images in my imageCollection. I don't want to create list and to type by myself all the numbers like this-

var listOfNumbers = [0,1,2,3,4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19];

2. How can I filter image by number of pixels so I don't get very small rasters?

My end goal is to calculate statistic for my polygon for many years and for that I can't have the small rasters or the ones that look like cheese with many holes due to the cloud mask.

Here is one of my codes-

/**
 * Function to mask clouds based on the pixel_qa band of Landsat SR data.
 * @param {ee.Image} image Input Landsat SR image
 * @return {ee.Image} Cloudmasked Landsat image
 */
var cloudMaskL457 = function(image) {
  var qa = image.select('pixel_qa');
  // If the cloud bit (5) is set and the cloud confidence (7) is high
  // or the cloud shadow bit is set (3), then it's a bad pixel.
  var cloud = qa.bitwiseAnd(1 << 5)
                  .and(qa.bitwiseAnd(1 << 7))
                  .or(qa.bitwiseAnd(1 << 3));
  // Remove edge pixels that don't occur in all bands
  var mask2 = image.mask().reduce(ee.Reducer.min());
  return image.updateMask(cloud.not()).updateMask(mask2).divide(10000)
  .copyProperties(image, ['system:time_start']);
};

//Create LANDSAT7 dataset

var dataset = ee.ImageCollection('LANDSAT/LE07/C01/T1_SR')
                  .filterDate('1999-01-01', '2013-04-29')
                  .select('B1','B2','B3','B4','pixel_qa')
                  .filterBounds(geometry)
                  .map(cloudMaskL457);




//RGB visualization

var visParams = {
  bands: ['B3', 'B2', 'B1'],
  min: 0,
  max: 1,
  gamma: 1.4,
};


//clip the dataset according to the geometry
var clippedCol=dataset.map(function(im){ 
  return im.clip(geometry);
});

// Get the number of images.
var count = dataset.size();
print('Count: ',count);
//print(clippedCol);
print(dataset,'dataset');

//function to calculate NDWI in LANDSAT7
var addNDVI = function(image) {
  var NDVI = image.normalizedDifference(['B4', 'B3'])
  .rename('NDVI')
  .copyProperties(image,['system:time_start']);
  return image.addBands(NDVI);

};

//NDWI to the clipped image collection
var withNDVI = clippedCol.map(addNDVI).select('NDVI');

var NDVIcolor = {
  min: -1,
  max:1,
  palette: ['FFFFFF', 'CE7E45', 'DF923D', 'F1B555', 'FCD163', '99B718', '74A901',
    '66A000', '529400', '3E8601', '207401', '056201', '004C00', '023B01',
    '012E01', '011D01', '011301'],
};
print(ui.Chart.image.series(withNDVI, geometry, ee.Reducer.mean(), 30));
print(ui.Chart.image.series(withNDVI, geometry, ee.Reducer.max(), 30));
print(ui.Chart.image.series(withNDVI, geometry, ee.Reducer.min(), 30));
print(ui.Chart.image.series(withNDVI, geometry, ee.Reducer.stdDev(), 30));

Answer 1

You are on the right track;

The calculation you already created should be build into a function that maps over the entire image collection, when you add the result as an image property you can then filter on it:

var ndviWithCount = withNDVI.map(function(image){
  var countpixels = ee.Number(image.reduceRegion({
  reducer: ee.Reducer.count(),
  geometry: geometry,
  crs: 'EPSG:4326',
  scale: 30,
  }).get('NDVI'));

  return image.set('count', countpixels)
})

print(ndviWithCount, 'ndviWithCount')

//filter between a range
var filter = ndviWithCount.filter(ee.Filter.rangeContains(
          'count', 0, 10))
print(filter, 'filtered')

var max = ndviWithCount.reduceColumns(ee.Reducer.max(),  ["count"])
print(max.get('max'))