Powered by Face Detection Auto Focus (FDAF), the first-of-its-kind technology in the world.

OVERVIEW


using faces to relate to the world

FEATURES


face detection in a nutshell

INTEGRATION


and performance specifics

OVERVIEW

blue_small_bar

using faces to relate to the world

FEATURES

blue_small_bar

face detection in a nutshell

INTEGRATION

blue_small_bar

and performance specifics

OVERVIEW

  • NAME: LifeFocus
  • USAGE: fast and accurate focus on faces
  • HOW: it uses analysis and registration tehniques
  • IP: LifeFocus is a patented technology
  • PART OF: DigitalAperture product family
  • TYPE: computer vision product, user experience technology
  • INTEGRATION: software and hardware (via IPU)
  • MARKETS: 1 2 3 4  7  5 8 9 10 11 12

Photography is all about capturing life, which happens while you are busy trying to focus. Focus is the key element behind every picture, as without it, photography would have no value.

FotoNation’s FaceDetection & Tracking technologies, already found in most tier-one smartphones, enable face detection auto focus (FDAF). At the core of LifeFocus, FDAF is the fastest and sharpest AF solution when faces are detected. It complements phase detection auto focus (PDAF), which can encounter challenges in low-light situations and in cases where the object of interest is at a greater distance.

Most of the pictures taken with a smartphone are snapped in imperfect lightning conditions, affecting overall image quality. LifeFocus can overcome this, delivering fast and accurate focus even in very low or very bright light conditions where other technologies struggle.

FEATURES

Enables 3-level hybrid autofocus solutions for next-generation imaging.

 

life-focus-feature

FDAF is the first software-only focus technology in the smartphone market that uses predictive AF working in 4 dimensions (full frame + time). It is the fastest and the sharpest AF solution when faces are detected. It also complements phase detection auto focus (PDAF) and laser detection auto focus (LDAF), which can encounter difficulties in low-light situations or when the object of interest is at a greater distance.

The reality is that most pictures taken with a smartphone are snapped in imperfect lighting conditions, thus affecting the overall quality. FDAF can deliver fast and accurate focus, even in some low-light conditions ranging from 3 EV (20 lux) down to 1 EV (5 Lux).

The image quality achieved using FDAF is similar to contrast-based auto focus (CDAF), currently regarded as the highest standard of imaging quality. More recent systems such as PDAF and LDAF also reduce focus time, but fail to achieve the same level of image quality.

Market-proven technology.

PDAF and LDAF compatibility.

Amazing low light quality.

Smooth system integration.

INTEGRATION

AF performance is influenced by sag orientation, variation with temperature, humidity, or variations in a batch of camera modules. In order to target and actually achieve 1-step focus, thus eliminating “the hunting” process, we need to be able to do an ample system characterization and calibration. After this process, our Face Detection and Accurate Eye Tracking technologies can be fine-tuned to compensate for all of these factors.

Production line processes (based on system characterization and calibration)  enable highly optimized algorithms.

Integrating LifeFocus technology into a camera module offers an unprecedented ability to maintain focus while following a subject, especially when faces are present. And this includes following a subject even in extremely difficult conditions like face obstructions, poor illumination, face angles or color casts.

life-focus-integration

Our characterization and calibration procedures can drastically improve the error rate of infinity and macro steps.

PRECISION – precise lens system positioning; no oscillations implies reduced power consumption.
SPEED – uses accelerated step and retains smooth convergence, thus generating less stress on the lens actuator.
ROBUSTNESS – tolerant to lens system placement impressions or image noise.