This weekend I read this book on augmented reality (AR): Augmented Reality A Practical Guide By Stephen Cawood, Dr. Mark Fiala First Edition January 2008 Publisher: Pragmatic Bookshelf Pages: 328 ISBN 10: 1-934356-03-4 | ISBN 13: 9781934356036 I learned a lot about the need for markers for synchronizing virtual reality with the real world. I learned how to set up some desktop AR demos using a compute This weekend I read this book on augmented reality (AR):Augmented Reality A Practical Guide By Stephen Cawood, Dr. Mark Fiala First Edition January 2008 Publisher: Pragmatic Bookshelf Pages: 328 ISBN 10: 1-934356-03-4 | ISBN 13: 9781934356036 I learned a lot about the need for markers for synchronizing virtual reality with the real world. I learned how to set up some desktop AR demos using a computer, Fiala’s ARTag system and a webcam. I learned the nitty-gritty of doing AR programming in C++ with Fiala’s ARTag API, Intel’s OpenCV computer vision library, and OpenGL graphics.I didn’t like the book much, however: it felt to me like an academic paper blown up to book length to no good purpose. I know that won’t make the authors or editors happy, but that’s my opinion.What I really want to figure out is how to build a handheld AR system based on a cell phone or digital camera. What spurred me on was a slightly cheesy episode of Numb3rs called Primacy, in which AR is used to extend a MMPRPG into the real world. I suppose that this would be possible to do with an iPhone using its SDK, although I’d worry about the iPhone’s limited bandwidth outside of WiFi hotspots. It might also be possible to do with an Android phone, once one exists. It could probably be done with other existing smart phones and even ordinary phones using their C++, C# and Java SDKs. It would be fairly easy using the .NET Compact Framework. I think I could do it using the BREW SDK, which I’ve used before, or the Symbian SDK.I’m not at all sure how much AR functionality I could shoehorn into the limited memory and CPU power of a garden variety cellphone, or how effective marker recognition would be with the limited resolution and slow response time of a standard cellphone camera.I’d be happy to hear from anyone who has experience in this area. Software Development