Researchers at Penn State have created a new device that mimics the neural network and red, green, and blue photoreceptors found in human eyes to produce images, drawing inspiration from nature.
“We acquired a plan from nature — our retinas contain cone cells that are delicate to red, green and blue light and a brain network that starts handling how the situation is playing out even before the data is sent to our mind,” said Kai Wang, partner research teacher in the Division of Materials Science and Designing at Penn State. ” This regular cycle makes the vivid world we can see.”
The researchers devised a novel sensor array made of narrowband perovskite photodetectors, which resemble our cone cells, and connected it to a neuromorphic algorithm, which resembles our neural network, to process the data and produce high-fidelity images in an artificial device.
“We took a design from nature—our retinas contain cone cells that are sensitive to red, green, and blue light, as well as a neural network that begins processing what we see even before the information is transmitted to our brain,”
Kai Wang, assistant research professor in the Department of Materials Science and Engineering at Penn State.
Photodetectors are necessary for cameras and numerous other optical technologies because they convert light energy into electrical signals. Narrowband photodetectors can shine on individual pieces of the light range, similar to the reds, greens and blues that make up noticeable light, the researchers said.
Wang stated, “In this work, we found a novel method to design a perovskite material that is only sensitive to one wavelength of light.” We made three different perovskite materials, and they are planned such that they must be delicate to red, green or blue tones.”
According to the scientists, the technology may provide a means of avoiding the use of filters found in modern cameras, which reduce resolution while also increasing the cost and complexity of manufacturing.
Silicon photodetectors in cameras assimilate light yet don’t recognize colors. The reds, greens, and blues are separated by an external filter, which only lets one color into each section of the light sensor, wasting two-thirds of the incoming light.
“Our design makes it possible to avoid some information loss when the light is filtered. As a result, we propose that this work may represent a potential future camera sensing method for enhancing spatial resolution.
Additionally, the scientists stated that the utilization of perovskite materials results in the generation of power as the new devices absorb light, potentially paving the way for battery-free camera technology.
Postdoctoral researcher Luyao Zheng said, “The device structure is similar to solar cells that use light to generate electricity.” It will produce an electric current when you shine a light on it. Therefore, unlike our eyes, we do not need to use energy to gather this data from light.”
This exploration could likewise set off additional advancements in fake retina biotechnology. According to the scientists, devices based on this technology could one day replace damaged or dead cells in our eyes to restore vision.
The researchers wrote in the journal Science Advances that their findings represent a number of fundamental breakthroughs in the development of perovskite narrowband photodetection devices—from materials synthesis to device design to system innovation.
Electron-hole pairs are formed when light hits perovskites, which are semiconductors. An electrical current is made by sending these electrons and holes in opposite directions.
The researchers created thin-film perovskites with electron-hole transport that is severely unbalanced, indicating that the holes move through the material more quickly than the electrons do. The researchers discovered that they were able to harness properties and transform the materials into narrowband photodetectors by altering the architecture of the unbalanced perovskites or the order in which the layers are stacked.
Using these materials, they built a sensor array and projected an image through it with a projector. A three-sub-layer neuromorphic algorithm was used to feed information from the red, green, and blue layers into signal processing and image reconstruction. A type of computing technology called neuromorphic algorithms tries to mimic how the human brain works.
Wang stated, “We tried different ways to process the data.” The picture wasn’t very clear when we tried to combine the signals from the three color layers directly. However, the image is much more faithful to the original when we perform this neuromorphic processing.
According to the researchers, the fact that the algorithm mimics the neural network found in human retinas could shed light on the significance of these neural networks for our vision.
Wang stated, “We can demonstrate that the neural network functionality is really important in the vision processing in human eyes by joining our device and this algorithm together.”
Additionally adding to this examination from Penn State were: Junde Li, a doctoral candidate and Swaroop Ghosh, an associate professor in the Department of Electrical Engineering and Computer Science; what’s more, Dong Yang, associate exploration teacher, Jungjin Yoon and Tao Ye, postdoctoral scientists, and Monastery Marie Knoepfel and Yuchen Hou, doctoral applicants, in the Division of Materials Science and Designing. Additionally, Shashank Priya, a professor of materials science and engineering and former associate vice president for research and director of strategic initiatives, contributed.
More information: Yuchen Hou et al, Retina-inspired narrowband perovskite sensor array for panchromatic imaging, Science Advances (2023). DOI: 10.1126/sciadv.ade2338
