BACKGROUND
Digital cameras have focusable lenses usable to capture sharp images that accurately represent the details within a scene. Some of these cameras provide manual focus controls. Many cameras, however, such those as in wireless computing devices (e.g., smartphones and tablets) use automatic focus (autofocus or AF) algorithms to relieve the user of the burden of having to manually focus the camera for each scene.
Existing autofocus technologies capture an image, estimate the sharpness of the captured image, adjust the focus accordingly, capture another image, and so on. This process may be repeated for several iterations. The final, sharpest image is stored and/or displayed to the user. As a consequence, autofocus procedures take time, and during that time the scene may have moved, or the sharpness may be difficult to estimate given the current scene conditions.
A stereo camera, such as a smartphone with two or more image capture components, can simultaneously capture multiple images, one with each image capture component. The stereo camera or a display device can then combine these images in some fashion to create or simulate a three-dimensional (3D), stereoscopic image. But, existing autofocus techniques do not perform well on stereo cameras. In addition to the delays associated with iterative autofocus, if each individual image capture component carries out an autofocus procedure independently, the individual image capture components may end up with incompatible focuses. As a result, the stereoscopic image may be blurry.
