Not sure if the thread is right for the forum section, but anyway. I was curious was happened to the state of solutions for the seemingly simple task of QR recognition that's surely frequently used in various games. Over the course of two past weeks I have tried seven different solutions, with all of them either being broken or having major drawbacks, a stark contrast with 2012-2013 where people got it to work easily if cached threads are to believe. The thing I'm trying to do is very simple: getting a string out of a QR codes in a video feed (I'm making a basic location-based quest where people travel to certain places in the city, find QR prints there and use the app to verify they found the right location). I have two iOS devices (4th generation iPad and iPhone 5S) with iOS 8.1.3 I tested all the following solutions on. To recap, here are ways to recognize a QR code from a video feed on a mobile device I'm aware of: ZXing library hooked to WebCamTexture video feed. Found here. Works in the editor, silently fails to recognize anything on an iOS device, no errors to work with. ZXing library hooked to Image feed from CameraDevice instance in Vuforia. Works in the editor, silently fails to recognize anything on an iOS device, no errors to work with. Antares QR package from Asset Store. According to reviews, abandoned a long time ago Simple QR package from Asset Store. Works perfectly in the editor, crashes the application on iOS after a few seconds, silently fails to recognize anything during those seconds. Uploading QR images into Vuforia Target Manager, exporting them into a trackable library, using that trackable library on Vuforia AR camera, never actually using augmentation functionality itself, hooking up to trackable states to determine which image was found in the video feed. Horrifyingly limited approach: only a small number of images can be included in the application, every activation of AR camera incurs instantiation cost for complex trackable object created for every marker from the database, most of the functionality wasted for such a simple task. Cloud recognition with Vuforia: removes the issue of image library size, but completely unfit for my use case both due to cost and due to the quest taking place in the areas like the metro system where absolute majority of users have no internet connectivity. Using Metaio QR recognition functionality. Two hundreds megabytes in library size are a deal breaker when I'm not using anything but a tiny QR recognition portion of that library. Metaio also requires manual changes to XCode project which is unacceptable as it makes it incompatible with Unity Cloud Build. Can someone recommend an actual working solution? All I'm finding is unresolved threads about ZXing from more than two years ago, abandoned blog posts and abandoned broken examples, so save from going with incredibly inconvenient QR-as-AR-marker route, I'm stuck. Any help greatly appreciated.