Photogeist is the augmented reality sucessor to Save 'Em, designed and implemented under the guidance of Dr. Ehud Sharlin at the University of Calgary. Like Save 'Em, Photogeist was written in C++ and uses the ARToolkitPlus library to provide visual tracking.
In Photogeist you, the player, assume the role of an amateur paranormal researcher. As a young child, you witnessed a floating phantasm, and ever since, you've made it your life's work to expose the ghosts living among us. Sadly, your enthusiasm exceeds your evidence, and your many letters to the editor, blog postings, and self-published newspapers have been dismissed as the ravings of a madman. But that's all about to change. Late one night, while poring over your correspondence, you notice an anonymous tip, which tells of a haunted laboratory not far from your apartment! Grabbing your trusty camera from the shelf, you run out the door. Your goal for tonight is simple: Document this lab with as many ghost-filled photographs as you can take, and use this evidence to prove the existence of these otherworldly beings once and for all!
Photogeist is played using a tablet PC, with an attached web camera. This entire setup acts as your "camera". The web camera sits on the back of your tablet and acts as the "lens". You, the player, look through the LCD screen of the tablet, through which you will see virutal ghosts, apparently floating before you. Your goal is to take pictures of these ghosts. But not just any pictures, mind you. Your goal is to take the best pictures you can, since the pictures you take are judged based on a variety of aesthetic properties (centering, size, etc.)
Of course, this isn't as easy as it sounds - your ghostly subjects are constantly in motion, and what's more, they're aware of your movements. As you get closer and closer to your prey, the ghosts will start to act erratically in an attempt to avoid your lens. Players will have to react smartly to the movements of the ghosts if they hope to get the best shots.
I will be candid; I am not as proud of Photogeist as I am of Save 'Em. I think the philosophy behind the game is solid - there is value in allowing players to physically move while they play, and I believe that using photography as a gameplay mechanic has a lot of potential. Unfortunately, there were a variety of problems which prevented my final implementation from being as good as it could possibly be. Some of these problems were technological in nature. For example, the marker-based visual tracking which I use is not really suited for quick movements; rapid movement induces motion blur in the video, which makes the tracking markers unrecognizable. But other problems are simply a result of a messy design - the game has no conditions under which the player can lose, and no real reward for "winning". As a result, there's no clear incentive for players to improve, and the game gets boring quickly.
Do not misunderstand me; I don't see Photogeist as a failure. From a technical standpoint, I think the game is incredibly interesting. But in the end, it just doesn't feel fun, and that is the metric by which all games must eventually be judged. I hope to take the lessons I've learned from Photogeist, and put them to good use in my future projects.