Aeternity Drone Graffiti Project

DroneTracer Library

The purpose of this library is to transform bitmap images into vector-based artwork flyable by drone.

The conversion is done in different stages in which the image is analyzed and the most relevant visual elements are extracted. A path, representing the image abstracted information, is calculated using path trajectories optimized for the drone parameters.

The flight time estimation and size/resolution is calculated along with the drawing coordinates.

The challenge is consistently transform images limited by the physical characteristics of the medium and perform these calculations using just client-base JavaScript.

API

Library name: DroneTracer

Painting Configuration:

Creating an instance of the DroneTracer class requires passing a configuration object. This configurations describes properties of the wall and the painting required by the transformation process and the drone system. Multiple instances of the DroneTracer class can be created simultaneously with the same or different configuration.

// Painting configuration
{
  wallId: 'CT19-001',		// required | ID to identify the wall target
  gpsLocation: [0,0],		// required | wall GPS coordinates [Latitude, Longitude]
  wallSize: [33000, 50000],	// required | wall width [mm], wall height [mm] (33m x 50m)
  canvasSize: [33000, 10000],	// required | canvas area size [mm] (33m x 10m)
  canvasPosition: [0,10000],	// required | relative position of the canvas in the wall [mm]
  colors: ['#000000'],		// list of available colors
  strokeWeight: 100,		// drone paint stroke thickness [mm]
  droneResolution: 200,		// drone resolution [mm]
  dronePrecisionError: 150,	// drone positioning error margin [mm]
  droneFlyingSpeed: 0.6,	// average drone flying speed [m/s]
  droneFlightTime: 240000,	// duration of battery flying [ms]
  droneDrawingTime: 84000,	// average continuous drawing time [ms]
  droneSwapTime: 300000,	// land, swap battery and paint can, takeoff, and resume painting [ms]
  droneTakeoffTime: 140000,	// max duration from drone takeoff to actual painting [ms]
  droneLandingTime: 90000,	// max time needed to stop painting and land [ms]
  minimumImageSize: [350,350], // Min image size to be accepted
}

Required parameters: wallId, gpsLocation, wallSize, canvasSize, canvasPosition

DroneTracer - transform( File, [progressFunction], [transformationOptions] )

Converts an input image into a svg drone paint

Parameters:

File: The input accepts a Base64 string or a File API object. File formats must be jpg and png.

Progress Function: |Optional| Function callback that return the progress of the transformation process. [from 0 to 1]

TransformationOptions object: |Optional| Object that adjusts transformation parameters.

Return: Promise The transform function returns a promise that resolves into the DronePaint object.

DroneTracer - uiParameters

The uiParameters getter, exposes the properly parameters, values and ranges for a successful image transformation.

This parameters can be used to build an user interface and allow the user to retouch the results.

[{
    label: "Blur Radius",
    name: "blurKernel",
    value: 4,
    type: "range",
    from: 1,
    to: 8
},
{
    label: "Threshold",
    name: "hysteresisHighThreshold",
    value: 50,
    type: "range",
    from: 1,
    to: 100
}]

dronePaint object:

Provides functions to access and modify related information to the svg for the drone. This functions does not affect the transformation process.

setPaintingPosition(x, y): Sets relative positioning inside the canvas area. |return false if position is out of bound|

setPaintingScale(factor): Transforms the scale of the generated SVG. Only accepts positive numbers greater or equals than 1. |return false if is scale is out of bound|

setPaintingColor(color): Sets the painting color of the draw.

svgFile: Getter, svg drone path

sourceImage: Getter, original source image. File API

estimatedTime: Getter, estimated painting time (milliseconds). Includes takeoff, painting and flying time, battery and spray can swapping and landing.

Implementation

// Painting wall configuration
var paintingConfig = {
    wallId: 'LN21-011', 		// required
    gpsLocation: [0000,0000], 	// required
    wallSize: [22000, 40000],	// required | millimeters
    canvasSize: [20000, 20000], // required | size of target canvas 
    canvasPosition: [0, 0], 	// required | millimeters (origin = [bottom left])
    colors: ['#000000', '#eb340f', '#0f71eb'], // default [#000]
}

// Instance of a drone tracer
var tracer = new DroneTracer(paintingConfig)

// get the transformation parameters to build the UI
var uiParameters = tracer.uiParameters

// Transform image into a flyable drone path
tracer.transform(
  imagefile, // loaded File API or Base64 String
  (progress) => { console.log(`${progress}%`) }, // log progress
  { 
    centerline: false
    blurKernel: 3,
    hysteresisHighThreshold: 50,
    binaryThreshold: 45,
    dilationRadius: 6     
  }
).then( (dronePaint) => {
  /* The dronePaint object, provides functions to access and modify
   * related information to the svg for the drone.
   * This functions does not affect the transformation process.
   */
  dronePaint.setPaintingPosition(12000, 0) // in millimeters
  dronePaint.setPaintingScale(2.5) // post-scale the svg
  dronePaint.setPaintingColor('#aa0000')

  console.log( 'result path: ', dronePaint.svgFile )
  console.log( 'image source: ', dronePaint.sourceImage )
  console.log( 'painting time: ', dronePaint.estimatedTime )
});

Development

Installation

$ npm install

Run test

$ npm test

Build Release version

$ npm run rlease

Test Interface Version

# Requires release version previously built
$ npm run interface

License

MIT

nicotron / DroneTracer