New AR Ipad game "Driver" by Ubisoft

A new AR IPad2 game "Driver" was just released by Ubisoft.

The gameplay seems to be kind of particular. The player basically steers the car around obstacles by way of pressing 4 different buttons (forward, backward, left, right) located on the left and right side of the tablet. Not the most innovative interface. Looks like a speed control technique.

Concerning the tracking technology, I would suppose they use Qualcomms QCAR SDK. Hopefully they will release more AR games in the near future.

Actually, you could imagine to use some kind of parallel tracking and mapping algorithm since the entire game happens on a planar ground surface, like the floor. This would avoid the need of a marker.




 

Talk on mobile Augmented Reality at Paris Android User Group (Paug)

Hi there,

here is the video of our mobile augmented reality talk given at Paris Android User Group in june.

In the first part Atef Haouari presented a technique to create a mobile AR application for Android using position (GPS) and orientation sensors. There is a recap of the entire talk here.

In the second part (from 1:07h), I present some of our work in the area of vision based mobile augmented reality for Android using the Qualcomm AR SDK (QCAR) and Unity3D.

The presentation was held in french. Sorry for all the non-french speakers.

Watch live streaming video from parisandroidug at livestream.com

Qualcomm AR SDK for iOS (IPhone & IPad)

Hi there,

just got my hands on the new Qualcomm AR SDK for IPhone and IPad, which you can download here. I tested the plugin Unity3D 3.4 and it works as smooth and easy as the earlier released Android SDK . There is also an SDK for XCode available, but I didn't test it yet. It's just so much easier to do some initial prototyping with Unity.

On a first glance it is very much the same as the Android SDK. There is the ARCamera object and the image targets. So if you familiar with the Android release you could probably port your project very fast to iOS. Which is great news.

An initial test (image right) using the standard Teapot example with two markers runs perfectly on IPad2. I did not measure the framerate for the moment, but I think it is far from touching the IPad2 resource capacities.

To get started Download the QCAR Unity plugin for Unity3D (you need the most recent Unity3d 3.4!). Unzip, install the included app and it will extract multiple Unity package files among you will find QCAR-ImageTargets.unitypackage. Load (just double click) this package into Unity3d and choose the delivered test scene. It contains two targets, each visualizing a classic Teapot. Deploy for iOS. Unity3D will create now the corresponding XCode project, which you can compile in XCode and deploy on your IPhone or IPad.

Have fun

Breizhcamp Recap

Hi there,

Last friday I went to Rennes to give a talk at the Breizhcamp on mobile augmented reality for Android using the Qualcomm AR SDK and Unity3D. Bassically an extended version of my previous PAUG talk in Paris. Slides are below. There will be as well a video I will post as soon as I get my hands on it.

Anyway, Breizhcamp was really cool. A lot of nice presentations around GWT, .Net, Android, iOS etc. There was also a talk about Kinect, which was interesting and quite entertaining at the same time.

Breizhcamp Rennes 2011

Slides of my mobile AR talk at PAUG now online

Interview at Droidcon 2011 in Berlin

My interview (in german!) with Golem.de at Droidcon in Berlin march 2011 on mobile Augmented Reality. Link to the article Golem.de http://www.golem.de/1103/82345.html

Initial thoughts on mobile AR (SDKs)

Novel augmented reality SDKs allow developers to create a new generation of innovative augmented reality games and applications for mobile devices that use Android. This new technology is able to estimate the handhelds’ pose from almost any desired 2D image or photograph.

Thanks to increasing performance of modern mobile devices, innovative operating systems and powerful third-­‐party libraries, developers are able to create a new generation of augmented reality applications. Such apps are able to locate their position and orientation (pose) in space, using cutting edge computer vision algorithms. This way, virtual information can easily be integrated into a real environment in real time, which is known as mobile Augmented Reality (AR).

One of the first mobile AR applications was Layar. It used integrated sensors, like GPS, compass and accelerometer, to determine the device approximate position and orientation. In the following, virtual information about the environment was integrated in the captured video image.

Another approach to augment reality uses computer vision algorithms. Instead of using sensors, it determines the cameras’ pose using a visual reference, which is usually a 2D image (marker). One popular application that proposes such an approach was ARToolkit. ARToolKit was early ported to Android (NyArtoolKit), but has never been really successful due to bad performance and the inflexible marker design.

Today, a novel generation of computer vision algorithms are able to estimate cameras’ pose from natural features inside the environment in real time. This means, that almost any 2D image or photograph can be used as a reference, if it contains a minimum amount of visual information (features). Therefore, almost any 2D image can be used as a reference marker to allow the tracking algorithm to extract the camera pose in real time on a mobile phone.

In October 2010, Qualcomm released a free SDK (QCAR) for Android powered devices, which uses such advanced visual tracking algorithms. The SDK can be used in combination with the Android SDK or NDK. There exists also a plugin for Unity3D. Using these tools developers can create mobile augmented reality applications, like games.