Ultra thin arrayed cameras that are biologically inspired can take high resolution images.

Twenty years ago, digital cameras were seen as a niche, poor quality alternative to film. Today, you can buy a camera capable of producing a 100 megapixel image. While the race for more megapixels has slowed in recent years, a team at KAIST are developing something which will become increasingly more important in the future; smaller, thinner cameras.

University Researchers from KAIST

Ph.D. Candidate Kisoo Kim and Professor Ki-Hun Jeong © KAIST

The cameras design was inspired by the eyes of Xenos peckii – a type of paper wasp. The compound eyes of the wasp differ from most insects, in that they have many hundreds of photoreceptors per lens. This gives the wasp a high level of sensitivity to motion, alongside a wide angle of view while retaining a small focal length.

As a result of this, the eyes of Xenos peckii offer a superior image to most other insects – roughly 50 times better than the compound eyes of other arthropods, and since the eyelets are close, it is possible that the images from each can then be partially overlaid to improve the overall fidelity.

With this inspiration as the basis, the team at KAIST were able to develop a micro image sensor, stacked with a unique arrangement of inverted micro lenses, multilayered aperture arrays and gap spacers along a planar CMOS sensor.

The team states that the camera has a total track length (or TTL) of 700 micrometers. This  was achieved with the use of microlenses, which unlike conventional camera lenses offer a smaller focal length with minimal aberration.

As a result of their small focal length – in conjunction with the relatively small aperture – microlenses also offer a wider field of view with a deep depth of field.  The team note that the image array in their ultrathin camera provides a field of view of over 70°.

Scaled array of images

Image 2. Super-resolution imaging by array images.

As with most camera systems, the image resolution is bottlenecked by the diffraction limit of the optics. To bypass this limitation, the team uses a form of super-resolution imaging algorithm to enhance the fidelity of the image array that was initially captured.

The final result is a biologically inspired arrayed camera system capable of producing high resolution, high contrast images. Despite the current technological limitations, the cameras small total track length and impressive fidelity are a proof of concept for similar developments in the future – it’s likely we will see similar micro arrayed cameras implemented in surveillance and mobile systems moving forward.

Source: KAIST


Ultrathin Endoscope Captures Neurons Firing Deep in the Brain

Real-Time Captcha Technique Improves Biometric Authentication

Holography Approach Improves Heads up Displays for Planes and Cars

Sensors made with neural networks can recognize images in nanoseconds

0 0 votes
Article Rating