xview-filter-geojson.js

// filters the initial geojson "xView_train.geojson" into separate files matching each training tif
// use like `node index.js`

const fs = require('fs')
const _ = require('lodash')
const gs = require('geojson-stream')
const through2 = require('through2')
const log = require('single-line-log').stdout

const IMAGES = 3000

// match features with their corresponding image
function filterBy (image) {
  return through2({ objectMode: true }, function (feature, enc, callback) {
    if (+feature.properties.image_id.replace('.tif', '') === image) {
      this.push(feature)
    }
    callback()
  })
}

// counter for convenience
let features = 0
const count = through2({ objectMode: true }, function (feature, enc, callback) {
  features++
  log(features)
  this.push(feature)
  callback()
})

const inputStream = fs.createReadStream('xView_train.geojson')
  .pipe(gs.parse())
  .pipe(count)

inputStream.setMaxListeners(IMAGES)

// create write streams
_.range(IMAGES).forEach(image => {
    var w = fs.createWriteStream(`train_${image}.geojson`)
    inputStream.pipe(filterBy(image)).pipe(gs.stringify()).pipe(w)
})