Back to Search
Start Over
Single-shot recognition of orbital angular momentum from speckles with spatially multiplexed point detection (SMPD)
- Publication Year :
- 2023
-
Abstract
- The widely adopted optical information detection strategies nowadays can be roughly classified into spatial-resolve-based and time-sequence-based methods. The former is highly reliant on detectors with spatial resolution that works in a limited spectrum range to capture the intensity distribution of the object, whereas the latter employs single-pixel detectors (SPDs) with improved reliability and stability in detection at the expense of time consumption due to sequential intensity recording. However, there is a scarcity of research on using high-performance SPDs for single-shot object information detection. To capture and extract features of objects efficiently and economically, in this work, we propose a single-shot optical information-detecting method, tentatively named spatially multiplexed point detection (SMPD) technology, in which some SPDs are randomly distributed at different locations to acquire the information of the object. To validate the validity of the proposed method, we demonstrate high-fidelity recognition of orbital angular momentum (OAM) modes from speckle patterns generated by the transmission of two orthogonally polarized vortex beams through a multimode fiber. It is feasible to apply this approach in wide applications, such as image transmission based on multiplexed-OAM decoding and recognizing hand-written digits. Compared with traditional image recognition methods, the new approach yields a recognition accuracy of over 98% with an effective detection area that is only 0.02% of its counterparts. With further engineering, the proposed method may spur many exciting developments in OAM-based optical communication systems, image classification, object detections, and other optical information detections.
- Subjects :
- Physics - Optics
Subjects
Details
- Database :
- arXiv
- Publication Type :
- Report
- Accession number :
- edsarx.2310.04658
- Document Type :
- Working Paper