Perovskite sensor array mimics human retina for panchromatic imaging
The mammalian eye is a complex system made up of rods and cones that process the image data before sending it to the visual cortex of the brain via the optic nerve. Researchers at Penn State University created a system using perovskite RGB photodetectors (MAPbX3) and a neuromorphic algorithm to mimic this system as closely possible.
In imaging, panchromatic imaging means that you can enhance monochromatic channels (e.g. RGB channels are enhanced using panchromatic data (intensity and not frequency). This means that for the retina, the incoming light not only determines the different colors but also their intensity. This is what gives the Mark I eyeball its wide dynamic range. In this experiment layers of MAPbX3 perovskites (X being Cl or Br, or a combination thereof) were stacked to form RGB sensors.
Researchers noted that the output from these sensor layers could be processed using a pre-trained convolutional network to produce a panchromatic final image. This could then serve a variety of purposes. Researchers have noted that perovskite can be used to create new digital cameras and artificial retinas. However, the resolution of the layers is limited by their ability to scale, as is the longevity. One possibility is to power at least a part of the system with the energy collected by perovskite layer, similar to the proposed perovskite solar panels.
Source:
https://hackaday.com/2023/06/02/perovskite-sensor-array-emulates-human-retina-for-panchromatic-imaging/
