Researchers develop system for improved latent fingerprint recognition

Newswise — Recently, a group of researchers led by Professor LONG Shibing from the University of Science and Technology of China (USTC) of the Chinese Academy of Sciences, collaborates with Prof. LIU Qi from Fudan University, developed an in-sensor reservoir computer program for latent fingerprint recognition with deep ultraviolet photonic synapses and a memristor array. This study was published in Natural Communication.

Deep ultraviolet (DUV) photodetectors play an important role in deep space exploration, environmental monitoring, and bio-information detection. However, conventional ex-situ DUV fingerprint recognition systems use a separate sensor, memory, and processor, which greatly increases the decision-making delay as well as the overall computing power. Inspired by the human vision system, the research group developed a DUV system for the RC sensor with optical synapses as the reservoir input layer and memristor device array as the learning network, which can sense and process in parallel to ensure high efficiency. and low energy consumption.

Lire Aussi :  She built a startup that raised millions, shares 3 tips

The research team used Ga-rich component design and developed amorphous GaOx (a-GaOx) photo-synapses with improved photoconductivity (PPC) effects. The non-linear mapping relationship of the DUV system of the RC sensor is constructed by applying 4-bit equal light pulses to the simulation so that the image pixel sequence information is sampled by feature values.

Lire Aussi :  Luis Ruelas accused of giving Teresa Giudice's daughters fake Cartier

Finally, pool output training was achieved with the stable polymorphic modulation characteristics of the memristor device array, allowing for the detection of DUV fingerprints on a small scale. The best accuracy of DUV fingerprint image recognition when using the two-factor strategy and this hardware system is almost identical to the simulated results. The system achieves 100% recognition accuracy after 100 training sessions and maintains 90% accuracy even in the presence of 15% background noise, in accordance with the anti-noise properties of DUV light.

Lire Aussi :  Short Interest in T-Mobile US, Inc. (NASDAQ:TMUS) Expands By 19.2%

This full-sensor DUV hardware RC system provides a good reference example for effective detection and secure use of latent fingerprints. It is also an important reference for developing smart optoelectronic devices in the DUV waveband.

“This prototype system … will provide more insight into the emerging sensor reservoir computing. Overall, the topic of this work is really interesting.” said one referee Natural Communication.


Leave a Reply

Your email address will not be published.

Related Articles

Back to top button