The modern cell phone has emerged as the one-for-all device, bringing web browsing, digital photography, media playback, and even basic word processing to a single handheld device. In fact, many of the built-in cameras easily rival the quality of standalone digital cameras. Phones are now available with autofocus and resolutions of up to 12MP — although such high resolutions do not come cheap. Before choosing a camera phone because of its great built-in camera, think carefully about what you’ll use the pictures for. If you’re just looking to take a few snaps and upload them to Facebook, you won’t need anything more than about 3MP of resolution. If you’re into photography, the chances are that you already have a decent digital camera, so you won’t need additional such capacity in your cell phone. In other words, few people need such extravagant built-in cameras, so many camera phones will fail to give good value for money.
Major manufacturers include Sharp, Nokia, Samsung, Motorola, Sony Ericsson, and LG Electronics. The resolution is typically in the megapixel range, although some very-basic models still feature VGA cameras (with a resolution of just 640 x 480 pixels).
The cameras of smartphones are used as input devices in numerous research projects and commercial applications. A commercially successful example is the use of QR Codes attached to physical objects. QR Codes can be sensed by the phone using its camera and provide an according link to related digital content, usually an URL. Another approach is using camera images to recognize objects. Content based image analysis is used to recognize physical objects such as advertisement posters to provide information about the object. Hybrid approaches use a combination of unobtrusive visual markers and image analysis. An example is to estimate the pose of the camera phone to create a real-time overlay for a 3D paper globe. On recent camera phones it is even feasible to provide an augmented reality overlay for 2D objects and to recognize multiple objects on the phone using a stripped down object recognition algorithm as well as using GPS and compass. Auto-geotagging can show where a picture is taken, promoting interactions and allowing a photo to be mapped with others for comparison.