David Robson brings the world of adaptive optics into sharp focus
It may have taken billions of years to evolve naturally, but after just 10 years of research scientists have artificially recreated a key component of the human eye that could help everything from cheap, portable cameras to expensive medical equipment to create crystal clear images.
The device is a liquid lens that can adapt its shape to focus on the scene it is viewing, in a way that is very similar to the human eye. Previous artificial lenses had relied on mechanical motion to move the optics physically in a way that focused the image, but the new lens uses no moving parts at all, making it more robust than its mechanical counterpart.
It’s an innovative development, but the liquid lens from Varoptic is just the latest pearl in a string of research that hopes to emulate the human eye’s ability to react seamlessly to changes in a scene. The movement is called adaptive optics, and 50 years of research finally seems to be culminating in practical, cost-effective devices that could benefit imaging and laser systems.
The lens consists of a drop of liquid and an electrode, with an insulating substance that separates the two components. When a potential difference is applied to the liquid, an electrostatic effect on the particles in the liquid causes the drop to either bulge or flatten, depending on the voltage applied. This change in its curvature alters the degree to which it focuses the light.
When calibrated and connected to a suitable control system it can automatically focus an image. ‘The first experiment was done 10 years ago in Grenoble, but we have taken this time to develop and validate the technology,’ says Bruno Berge, the founder of Varioptic and one of the original researchers.
One of the first problems the team faced was to find a suitable way of packaging the liquid that would be robust to shocks and varying temperatures. They eventually decided on a package similar to the tiny ‘button’ batteries found in portable equipment. ‘It’s the only other component that contains liquids sealed in a tight way that can sustain varying temperatures for many years,’ says Berge. In this case, it has a small glass window to transmit the image.
A drop of liquid changes its shape when a voltage is applied, opening up the possibility for adaptive lenses.
In addition to the liquid that forms the actual lens, the packaging also contains a second liquid that acts as a kind of buffer against shocks. A lot of the team’s work has been to find liquids of a similar density so they do not move about when the cell is shaken. The liquids also needed to have good optical properties and good stability under different temperatures, and they could not be toxic.
The lenses are smaller than most autofocus mechanical lenses, meaning that they are ideal for small, portable cameras in webcams, mobile phones and bar code scanners. The liquid lenses also consume less power: mechanical lenses work at 300mW, whereas Varioptic’s liquid lens can function at just 15mW. Berge hopes that the lenses may also be used in medical endoscopes, as the small package is completely sealed and could be cleaned more easily than other systems.
In the future, Berge hopes to produce a zoom lens that operates using a similar mechanism. This June it produced a lens, in collaboration with Fraunhofer Institute for Applied Optics and Precision Engineering, Jena, Germany, that can magnify images at 2.5x, although the work is still in its very early stages. ‘Today we are trying to make prototypes,’ says Berge. ‘There has been no decision to make a product at the moment.’
Deformable mirrors correct aberrations
Accommodating for changes in the distance of an image is just one problem an imaging system has to face when focusing images, with astronomers and doctors frequently confronting other kinds of aberrations that can create blurry images. And it’s not just in imaging applications: frequently beams emitted from a laser are distorted and need to be adjusted before they can be applied successfully.
The trouble is, these aberrations do not stand still – they change with time, and from subject to subject. Winds blow high in the Earth’s atmosphere to distort the path of light from far away stars; changes in the eye’s physiology blur retinal images; and heating effects in a laser cavity cause the beam to drift slowly out of focus with time.
To overcome this, scientists have created shapeshifting mirrors that adapt automatically to the various aberrations to provide crystal clear images and well-shaped beams. The idea is not new – it was first proposed in 1947 – but until recently the equipment was so expensive that it was only really viable for military and large-scale astronomical operations.
The equipment typically contains three core components – the mirror, a wavefront sensor and a control system – that form a closed loop. The wavefront sensor consists of an array of lenses that focus light from the image onto a grid of photon sensors. When an image experiences aberrations, the light does not hit the lenses head on, and it is not focused in a regular pattern on the grid of photo sensors.
From this pattern of focused light, the control centre determines where the distortions are occurring in the image, and this information is used to determine what shape the mirror will need to take to correct the aberrations. Tiny actuators mounted on the mirror then push and pull it into the required shape. The process is so quick it can correct errors in the image in real time.
Bruno Berge, CSO of Varioptic: ‘We have taken time to develop and validate the [liquid lens] technology’.
As recently as 10 years ago these systems would have cost in the region of e10m, making them unfeasible for most applications. However, in 2000, Chris Dainty, a member of the Applied Optics Group at the National University of Ireland, Galway, and a team of researchers set about finding ways to cut these costs to as little as €15k.
‘There had been a culture that believed you couldn’t [build these systems] without 10 to 15 engineers,’ he explains. ‘We said “it’s not rocket science” and demonstrated that a PhD student could do it in three months.’
The team took a rough-and-ready approach to building the adaptive optics by compromising some of the functionality for lower-costs. Most of these price cuts arose from using low-end components that may not be perfect for highly accurate observations, but would still be suitable for many applications. For example, CMOS sensors, used in the wavefront sensors, provided a considerable saving compared to CCD sensors.
‘We obtained 80 per cent of the performance, with a dramatic effect on the cost,’ says Dainty. ‘Seven years on and there is now a bigger community in low-cost adaptive optics, which means the systems are now more usable and greatly improved.’
In addition to being cheaper than ever before, adaptive optics are now more powerful too. New FPGA processors, which accelerate the analysis of results from the wavefront sensor, provide more rapid control of the deformable mirror. The mirrors themselves are also developing, with an increasing ability to create intricate deformations in the reflective surface that can correct more complicated aberrations.
The optics’ primary application is still in biological imaging, to capture clearer images of body tissue such as the eye’s retina, even through aberrations in the lens and cornea. However, Alexis Kudryashov, a specialist from Russian company Active Optics, believes that adaptive optics will soon be prevalent in laser systems to correct aberrations and improve beam quality.
During its production, the transition of the laser beam through many different media such as the laser crystal, various optics and an amplifier, can damage its beam quality, together with heating effects in the laser medium and turbulence in the air. Adaptive optics, however, can correct these aberrations, in real time. They can also be used to shape the intensity distribution of the beam into a finely focused spot, which can be used for delicate laser processing.
‘To shape the beam using traditional optics is only possible if the beam is constant,’ says Kudryashov. ‘It’s not very simple to predict these changes, so you need active control.’ However, despite the recent reductions in the price of these components, the cost may still be too high at the moment for most laser suppliers to consider integrating deformable mirrors into their systems.
Kudryashov believes that a lot of further investment will be necessary until they do become a ubiquitous tool in optical systems. The idea of adaptive optics has taken a long time to mature, and it may still be some time before their potential is fully realised. But it’s possible that this longevity is proof in itself that adaptive optics are not just a new fad that will be soon be replaced by newer, more trendy technology. It seems they are here to stay, and they will only improve with age.