Physicists from the group of Jorik van de Groep at the UvA-Institute of Physics have devised a new method that can be used to detect edges of images in an extremely energy efficient and ultrafast way. The results were recently published in the journal ACS Photonics.

Computing with light
The energy cost of computation and data processing has become one of the most significant energy problems for our society. It is set to keep increasing its footprint in the future, as software demands keep increasing and current hardware is unable to provide a reliable answer. As such, over the past years, research into alternative methods of computation which are ultrafast and require minimal energy consumption has drawn significant attention.
One method of alternative computation is known as optical analog computing. In this method, mathematical operations are performed using light, even before it is captured by a camera. Since optical analog computing devices are passive – that is, they do not need electrical power – almost no energy is consumed, and moreover, the operation is literally done at the speed of light.
Detecting edges
In their research, together with industrial partners WITec and SCIL Imprint Solutions, the physicists have focused on edge detection techniques, aimed at identifying edges in images – locations where a sudden change in brightness occurs, indicating the border of an object that is observed. Edge detection is one of the most crucial tasks in image processing with applications in e.g. autonomous vehicles. To perform the optical analog computing, the physicists used a simple and easy to fabricate stack of thin films. The method turned out to work very well, being able to detect the edges of even very small objects, about 1 micrometer in size. Bernardo Dias, first author of the publication, says: “The design of the layer stack is extremely simple compared to the complex optical coatings that pose as the state-of-the-art. Despite this, our device shows one of the largest numerical apertures to date, allowing us to perform edge detection on the smallest possible targets.”

An additional benefit of the method is that it can work with a large number of light sources like lamps, LEDs or lasers, facilitating its potential use in existing technology. The results demonstrate that these techniques can in particular be used for high resolution microscopy. Since the device also highlights the edges of transparent objects which would be invisible to a conventional bright field microscope – think of cells – application in biological samples is also possible. As a next step, the researchers aim to develop switchable devices for optical analog computing, where one can switch the mathematical operation on and off, or where the device can switch between different functions.

Publication
High-NA 2D Image Edge Detection Using Tamm Plasmon Polaritons in Few-Layer Stratified Media, Bernardo S. Dias and Jorik van de Groep, ACS Photonics 12 (2025) 1, 311–319.