Camera with compound eye-like lenses


180 microlenses form the working components of the camera.— Photo: University of Illinois andBeckman Institute

A digital camera that has a lens that very closely mimics the compound eye of arthropods in all respects — wide-angle field of view of nearly 160 degrees, low aberration, high sharpness of vision, and infinite depth of field — has been developed by a team of scientists led by Young Min Song from the University of Illinois at Urbana-Champaign, U.S.

A paper on the invention is published today (May 2) in Nature .

Compound eyes of arthropods are by default hemispherical and have multiple lenses; hence any camera lens that intends to mimic them should have the same shape and multiple lenses. Digital cameras available today have a planar sensor with a single lens.

The biggest challenge the scientists faced was in developing such a hemispherical sensor that has multiple microlenses. They took advantage of the recent developments in stretchable electronics to achieve this.

Elastic microlenses that could be blown into a dome-shaped structure are formed from a moulded piece of rubber. An array of 16 by 16 microlenses is found in a small square area of nearly 15 mm by 15 mm. According to the authors, of the 256 microlenses present, only 180 form the “working components of the camera.”

Each convex microlens is connected to the base layer by means of a supporting post. This makes a microlens to appear like a dome on top of a pillar. A perforated black matrix covers the interspaces between the microlenses to prevent any stray light from entering the imaging system.

A black flexible silicon base layer has photodetectors that are arranged in such a manner that they match the microlenses.

The two layers — one containing the lenses and the other containing the photodetectors — are then bonded in such a manner that the photodiodes are at the “focal position” of the lenses. The bonding is done at the points where the lenses overlie the photodetectors.

Since both layers are made of stretchable material and are bonded at the correct points, they can be elastically changed from a flat shape, in which they are fabricated, to a dome-shaped structure when it becomes a part of the camera. The dome-shaped structure of the lens mimics a compound eye.

The authors stress that changing the shape from planar to hemisphere neither changes the optical alignment nor the optical and electrical. performances.

The resulting hemisphere-shaped imaging system with multiple microlenses behaves like an apposition eye — compound eye found in diurnal insects.

The microlens, supporting post, photodiode and the black silicon base correspond to an eye’s corneal lens, cone, rhabdom (channel that contains light-sensitive cells) and screening pigment respectively. All the lab-created structures together form an optical unit (ommatidium). “Each ommatidium contributes a single pixel.”

When an object is placed within the viewing angle (about 160 degree) of the camera, the microlenses facing the object produce a small image of the object. The photodiodes corresponding to the microlenses that produce an image generate photocurrents.


The biggest concern is image distortion when the imaging system is changed from a planar to hemisphere shape. However, they found that it did not suffer from off-axis aberrations. The paper displays this through an illustration of soccer ball placed at three different angular positions — from left to centre to right of the centre of the camera. The object does not get distorted or blurred with a change in angular position.

The short focal length of each microlens produces an “infinite depth of field.” As a result, as an object moves away from the camera, size of the image is reduced but the image remains in focus.

Published in The Hindu on May 2, 2013