This innovative camera captures everything in focus simultaneously.
A camera lens has traditionally focused on just one subject at a time, much like our own eyes. However, that may soon change thanks to innovative lens technology from researchers at Carnegie Mellon University (CMU). This new development allows for every part of a scene to be in sharp focus, capturing intricate details across the entire image, regardless of distance.
Conventional lenses can only sharpen one focal plane at a time, resulting in blurriness for objects in front of or behind the focused subject. While this creates a sense of depth, achieving a clear overall image usually requires combining multiple shots taken at different focal lengths. The new “spatially-varying autofocus” system merges various technologies, enabling the camera to determine which areas of the image should be sharp—effectively giving each pixel its own tiny, adjustable lens, as explained by CMU associate professor Matthew O’Tool.
The researchers have created a “computational lens” that integrates a Lohmann lens—comprising two curved, cubic lenses that shift against each other to adjust focus—with a phase-only spatial light modulator, which controls how light bends at each pixel. This setup allows the system to focus at multiple depths simultaneously. It also employs two autofocus techniques: Contrast-Detection Autofocus (CDAF), which segments images into regions to independently maximize sharpness, and Phase-Detection Autofocus (PDAF), which identifies whether an object is in focus and determines the necessary adjustments.
The experimental system “could fundamentally change how cameras see the world,” according to CMU professor Aswin Sankaranarayanan. While this technology isn’t yet available in any commercial cameras, and it may take time before it hits the market, CMU researchers believe it could have wider applications beyond traditional photography. Potential uses include enhancing microscope efficiency, creating realistic depth perception for VR headsets, and providing autonomous vehicles with “unprecedented clarity” in their surroundings.