3D Sensing

3D Sensing

Although 3D sensing technologies have been in development for many decades, their adoption into high volume electronic devices has only recently been enabled by improvements in computing power and image sensors.

3D sensing underpins a range of features that include:

  • Gesture Recognition – the ability to control a device by moving hands, fingers or eyes.  3D information is required to decide the direction a finger is pointing for example, particularly if it is pointing towards the device.
  • Augmented Reality – real time images are overlaid with relevant information or virtual objects are inserted for the purpose of entertainment.  Implementations of augmented reality require the 3D position of objects in the image to be convincing.
  • Facial Recognition – a security measure that accurately identifies an individual and is not subverted by a photograph.  The key to discriminating between a real face and other objects is a comprehensive 3D representation.
  • Measurement – the ability to measure an object accurately and rapidly in 3D.  This has a variety of commercial uses such as pricing the delivery of an object or 3D modelling for the purpose of additive manufacture.
  • Object detection and recognition – the ability to recognise objects, hazards and obstacles.  This is required for autonomous devices to perform advanced functions such as delivering a package, or driving.

It is clear that 3D sensing technologies open an enormous range of applications across a very broad base of product lines in several categories.  It is expected that 3D sensing applications will see enormous growth and investment in the next few years.

Some 3D sensing technologies work in a similar way to the human eye.  Two images are made from different positions and the images are combined and the distances to various objects computed.

Competing technologies beam structured light onto an object and then make images of the light pattern to measure the surface.  The structured light is often infra-red and is a known pattern. The pattern will be distorted when it impinges on a surface and the distance to the points on the surface can then be calculated.

There are a number of ways that CML's patented technologies can improve the accuracy and performance of 3-D measurement systems:

  • If the dual camera method is used, better results will be obtained if the cameras are stabilised using OIS, particularly in low-light situations with longer image collection times.
  • If the structured light method is used, better resolution can be obtained by moving the light source or the image sensor.  CML's unique ability to move objects with sub-micron precision is a perfect fit for this application.