What to Expect When Choosing ArUco or ARToolKit

For developers prioritizing ease of use and effective object recognition, ArUco is your go-to choice. However, if interactivity within AR environments and powerful tracking capabilities take precedence, choose ARToolKit. Don’t overlook their specific project applications: ArUco excels in video surveillance, while ARToolKit maximizes utilization in education and geometry.

Key Distinctions Between ArUco and ARToolKit

• Developers: ArUco was developed by Rafael Munoz and Sergio Garrido, while Hirokazu Kato, Mark Billinghurst, and Ivan Poupyrev are behind ARToolKit.
• Object recognition: ArucO leverages a marker-based technique using black and white markers. ARToolKit tracks planar images and simple black squares.
• Adaptability: ArUco integrates seamlessly with OpenGL and OGRE, whereas ARToolKit is compatible with Unity and OpenSceneGraph plugins.
• Specialties: ArUco thrives in non-immersive AR applications, useful in video surveillance and obstacle avoidance. ARToolKit enhances interactivity within AR environments, best utilized in education and geometry realms.
• Platform support: Both technologies cater to different systems. ARToolKit supports Windows, OS X, Linux, iOS, Android whereas ArUco primarily works in conjunction with OpenCV.

What Is ArUco and Who’s It For?

ArUco, a condensed library conceived by Rafael Munoz and Sergio Garrido, brilliantly exploits OpenCV for augmented reality applications. Dominantly employed for camera calibration, ArUco’s prowess extends to object recognition, carefully molting processes like detection, processing, and identification. Intuitively developed with black and white square-shaped markers, ArUco shines in its marker-based approach for pose estimation, thereby enhancing object recognition. Whether it’s non-immersive AR applications, video surveillance, or static obstacle avoidance, ArUco delivers with aplomb. Anyone involved in multimedia enrichment, human posture evaluations, and interactive educational set-ups would find tremendous value with ArUco.

Pros of ArUco

• Effective object recognition through distinct marker-based techniques
• Integration with OpenGL and OGRE provides fast and efficient solutions
• Significant contribution to video surveillance and obstacle avoidance
• Promotes interactive education by boosting engagement and interactivity

Cons of ArUco

• Affordability issues might deter potential users
• High skill requirements could limit accessibility
• Lack of broad-scale application potential

What Is ARToolKit and Who’s It For?

Framed by Hirokazu Kato, Mark Billinghurst, and Ivan Poupyrev in 1999, ARToolKit stands as a testament to open-source, real-time augmented reality. With University of Washington HIT Lab as its birthplace, ARToolKit transitioned under DAQRI’s care before ARToolWorks v1.0 surfaced. Notable for its single-camera and stereo-camera tracking ability, ARToolKit finds its stronghold in domains spanning over education, spatial relations to molecule structures. If you seek a system that combines camera calibration, marker generation utilities with wide OS compatibility for augmented reality, ARToolKit is your go-to choice.

Pros of ARToolKit

• Inclusive OS support: Windows, OSX, Linux, iOS, Android
• Enables user interactivity with virtual objects overlaid on real camera feed
• Resolves tracking, registration, calibration complications for an easy development process
• Open-source software, broadening the audience reach

Cons of ARToolKit

• Slight technical knowledge required might limit the userbase
• Hardware specific requirements such as 500 MHz processor and a graphics accelerator card
• AR overlays depend on the display technology used

Code Examples for ArUco & ARToolKit

ArUco

Our first ArUco example plays with perspective transformation, an essential concept in augmented reality. But here’s the catch: we’re creating a virtual billboard that “orients” based on the viewer’s perspective. You’ll need OpenCV (for Python), numpy, and cv2.aruco. Make sure to include an image of your choice for the virtual billboard.

import cv2
import numpy as np
import cv2.aruco as aruco

def loadCoefficients(path):
    with np.load(path) as X:
        matrix, dist, _, _ =  for i in ('mtx','dist','rvecs','tvecs')]
    return matrix, dist

def getPoint3D(point2D):
    return np.array(, point2D, 0.0, 1.0])

cap = cv2.VideoCapture(0)
dictionary = cv2.aruco.getPredefinedDictionary(aruco.DICT_6X6_1000)
parameters = aruco.DetectorParameters_create()

while True:
    ret, frame = cap.read()
    gray = cv2.cvtColor(frame, cv2.COLOR_BGR2GRAY)
    corners, ids, rejectedImgPoints = aruco.detectMarkers(gray, dictionary, parameters=parameters)
    for corner in corners:
        cv2.polylines(frame, corner.astype(int), True, )
    cv2.imshow('window', frame)
    cv2.waitKey(1)
cap.release()
cv2.destroyAllWindows()

ARToolKit

Now, let’s move to the ARToolKit by coding an interactive AR tool; a flashlight that interacts with an AR object and virtual lights. This example requires AR2 and ARspot, part of ARToolKit. Without further ado, let’s jump in:

let scene, camera, renderer;
let ARlight, ARobject, ARspot;

init();

function init() {
    scene = new THREE.Scene();

    let spotLight = new THREE.SpotLight(0xffffff);   
    spotLight.position.set(100, 100, 100);
    scene.add(spotLight);

    ARlight = new THREE.PointLight(0xc4c4c4, 10);
    ARlight.position.set(0, 300, 500);
    scene.add(ARlight);

    ARspot = new THREE.SpotLight(0xffc95c,10);
    ARspot.decay = 2;
    ARspot.penumbra = 0.05;
    ARspot.angle = 0.35;
    ARspot.position.set(0, 30, 30);

    let spotLightHelper = new THREE.SpotLightHelper(ARspot);
    scene.add(spotLightHelper);

    ARlight.add(ARspot);

    ARobject = new THREE.Mesh(
        new THREE.SphereGeometry(5, 70, 70),
        new THREE.MeshStandardMaterial()
    );

    ARobject.position.set(0, 0, 25);
    scene.add(ARobject);

    renderer = new THREE.WebGLRenderer({
        antialias: true
    });

    renderer.setSize(window.innerWidth, window.innerHeight);
    document.body.appendChild(renderer.domElement);
    
    render();
}

function render() {
  ARlight.position.set(
        ARobject.position.x - 2,
        30,
        ARobject.position.z - 2
    );

  renderer.render(scene, camera);
  requestAnimationFrame(render);
}

The Final Claim: ArUco vs ARToolKit?

There’s no one-size-fits-all when it comes to AR libraries. Let’s dissect the options for different audience segments.

AR Devoted Educators

If you are focusing on creating enhanced learning experiences, ArUco provides a robust option, especially advantageous with its capability to create interactive multi-sensorial learning environments. It also simplifies object recognition processes, allowing for engaging, immersive lessons in various subjects, from human physiology to celestial navigation.

The AR Visionaries

The visionaries inventing the next AR experience should leverage the prowess of ARToolKit. With its expansive platform support including Windows, OS X, Linux, iOS, and Android, and the ability to maximize user interactivity in AR environments, you have the power to spark change universally.

The Surveillance Engineers

In the realm of video surveillance, ArUco stands as an unparalleled pick, thanks to its efficacy in detecting anomalies and tracking structures. Not only can it handle occlusion problems effectively, but its fast processing modes also ensure you won’t miss a beat when it comes to keeping areas secure.

AR Gaming Titans

For the gaming industry titans, ARToolKit offers a hard-to-beat proposition. Its compatibility with Unity plugins and support for head-mounted displays nudge it toward the finish line. Immerse your players like never before and create experiences they’ll never forget.

When it comes to choosing between ArUco and ARToolKit, it’s a matter of matching expertise to needs. ArUco pierces its way through object recognition tasks and suits education and surveillance scenarios perfectly. ARToolKit, conversely, offers a wide platform reach and interactivity making it an ideal choice for game developers and AR pioneers.