I have a collection of Landsat images filtered over forest regions of India over a time interval of 6 years(2013 - 2018). I have again filtered this collection such that I have one image per 4 months, that is I have filtered on a quarterly basis so that I will have 3 images per year.

Now, for a particular point, there will be 18 pixels representing it over 6 years. I have to find, out of these 18 pixels, how many satisfy the condition - NDVI < 0.25. Suppose, 10 pixels satisfy this condition, then in my output image, the pixel representing that particular point should have its value = 10. Like this I have to do for all the points in my selected region.

I have no idea as to how to do this pixel wise computation. Can somebody help me on this?

In my code, I have done till filtering quarterly wise.


var treeCanopyCover = hansen.clip(india).select('treecover2000');
var greater25 = treeCanopyCover.gte(25);


function addNDVI(image) {
var ndvi = image.normalizedDifference(['B5', 'B4']).rename('NDVI');
return image.addBands(ndvi);


var startDate = ee.Date('2013-01-01');
var endDate = ee.Date('2019-06-30');
var filtered = l8.filterBounds(india)
      .filterDate(startDate, endDate)

var maskedCollection = filtered.map(function(image){
  var x = image.updateMask(greater25);      // filtering over forest regions
  return x;


var map_m = function(y)
  var months = [1, 5, 9];
  var filtered_col = months.map(function(m) 
var start = ee.Date.fromYMD(y, m, 1)
var end = start.advance(3, 'month');
var filt = maskedCollection.filterDate(start,end).mean();

return filt.set('year', y)
          .set('month', m)
          .set('system:time_start',ee.Date.fromYMD(y,m,1)) ;
  return filtered_col;

var years = ee.List.sequence(2013, 2019);
var quarterlyimages = years.map(map_m).flatten();
var z = ee.ImageCollection(quarterlyimages);

1 Answer 1


I think count method of ImageCollection could help with what you're doing.

Assume col is the ImageCollection that you want to deal with. Every image in this col has NDVI band that holds NDVI value. The code below will result in an image called NDVI_count in which value of every pixel represents the number of pixels at the corresponding location in col that have NDVI < 0.25.

var NDVI_count = col.select('NDVI')
  .map(function(img) { return img.updateMask(img.select('NDVI').lt(0.25)) })

The map function in the above code just does one thing: mask out any pixels that have NDVI >= 0.25. The count method will count the number of valid pixels (i.e. not-masked pixel) throughout the collection at each specific location.

Hope this helps.

  • This gives me the number of all pixels with NDVI less than 0.25. I need the number of pixels specific to a particular point, say for a particular longitude and latitude. Jul 9, 2019 at 4:14
  • Sorry if I was not clear enough. It does give you the number of pixels with NDVI less than 0.25 at each particular location.
    – Kevin
    Jul 9, 2019 at 5:40
  • Yeah, it worked. Sorry for the misunderstanding. I didn't expect that such a small single inbuilt function could solve this. Thank you very very much Jul 9, 2019 at 9:41
  • Just one more question: If I have to plot a bar chart or histogram showing the frequency of each count, that is how many pixels in the output image have the value (say, 3), how can I do that? @Kevin Jul 9, 2019 at 9:52
  • 1
    @ReemaMathew, you can use ui.Chart.image.histogram to create such chart. To specify which chart type, setChartType can be used after creating the chart.
    – Kevin
    Jul 9, 2019 at 11:25

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.