Digital-optical computational imaging capable of end-point logic operations

Significance 

Engineering sciences have played a major role in advancing imaging by improving and miniaturizing detectors, enhancing system design, increasing speed, sensitivity and resolution, accelerating computational analysis, and developing methods to minimize the side effects of applied energy. Computational imaging has been identified as a new promising strategy. This framework provides different imaging modalities and its working principle is based on a combination of computational decoding and optical encoding. Through the encoding-decoding combination, it is much possible and easier to achieve high-functional imaging.

The field of computational imaging has two main distinctions: computational photography and optical computing. The former mainly uses light field cameras to capture light array signals, allowing object space reconstruction after image capturing. On the other hand, optical computing uses optical signal processing for light signal encoding and computational decoding to improve the imaging properties. Additionally, in this field, numerous methods, such as comprehensive sensing, have been developed based on signal theory for functional imaging-related applications. Lately, artificial intelligence and machine learning have become of great significance for improving the properties and capabilities of computational imaging, opening a route for new practical application possibilities.

Herein, Osaka University scientists, Professor Jun Tanida, Ms. Karin Tsuchida and Mr. Ryo Watanabe developed a new digital-optical computational imaging with high selectivity and end-point logic operations capabilities over free-space data transmission. This framework is considered an extension of computational imaging because it uses digital-optical codes that were originally designed for digital optical computing. Their research work is currently published in the journal, Optics Express.

In their approach, spatial code patterns designed for the optical logic operations were extended to digital optical codes in the spectral and temporal domains. These optical codes could be differentiated from the background signals due to their distinct artificial forms. The optical logic operation feature enabled the end-point logical operations during imaging. Additional binary functions were generated by combining the spatial codes encoded from the signal inputs. Data transmission performed by assuming the transmission of the signal from the sensor nodes to the device was utilized to demonstrate the features of this framework. Furthermore, the transfer of encrypted data was demonstrated, and the effectiveness of the proposed framework was clarified.

The research team showed that though the end-point logic was realized by the same method similar to that used in the original optical logic operator, various physical implementation options, including time, space and spectral domains, could still be selected to realize various requirements like enhanced data-transmission bandwidth. Consequently, by carefully and appropriately selecting the physical forms of the digital-optical codes, it was possible to improve the data transmission bandwidth by combining different forms selected.

Moreover, the transfer of encoded signals over free space followed by decoding on the destination device also allowed logical operations at the end-point of the data transmission. For the experiment, the encrypted signals were captured as image using the digital-optical codes, while the original signals were correctly retrieved using the end-point exclusive operation.

In summary, the study reported the successful design and demonstration of innovative digital-optical computational imaging for efficient object transmission with the ability to perform end-point logical operations. The capability of the proposed framework was extended by using the combination of the data-transfer modes used in assigning the preprocessing of signals to be encoded as well as the end-point processing. In a statement to Advances in Engineering, Professor Jun Tanida, the lead and corresponding author said that their study will extend the capabilities of digital-optical computational images for various new applications.

Reference

Tanida, J., Tsuchida, K., & Watanabe, R. (2021). Digital-optical computational imaging capable of end-point logic operationsOptics Express, 30(1), 210-221.

Go To Optics Express

Check Also

Wires as Heatsinks: Reducing Thermal Resistance in CMOS-SOI ICs - Advances in Engineering

Wires as Heatsinks: Reducing Thermal Resistance in CMOS-SOI ICs