image: Figure 1 | Schematic diagram of drone detection.
Credit: Shuxiao Wu et al.
The low-altitude economy, as an emerging economic form, has experienced rapid growth in recent years. Given the significant increase in the number of drones and the frequency of flight activities, the absence of effective regulation poses a substantial threat to aviation safety and public security. In this context, the development of drone detection technology assumes critical importance. However, traditional imaging technologies are constrained by insufficient imaging sensitivity and background noise interference, which hinder their ability to meet the practical demands of long-distance drone detection.
Single-photon imaging (SPI) technology has garnered significant attention due to its capability to markedly enhance imaging sensitivity and substantially extend imaging distance. However, in practical applications, drone detection often encounters complex environmental backgrounds (e.g., urban buildings, open skies, dense jungles, etc.), which renders the imaging process vulnerable to background noise interference. Extracting moving target features for imaging can effectively mitigate background noise interference and considerably improve the imaging signal-to-noise ratio, particularly for point targets such as drones, thereby providing a novel dimension for feature recognition. Nevertheless, traditional photon counting techniques typically necessitate lengthy integration times to compensate for intensity fluctuations caused by shot noise, leading to a reduced imaging frame rate and an inability to capture high-frequency dynamic characteristics of the target object.
In a newly published paper in Light: Science & Applications, a team of scientists, led by Professor Liantuan Xiao and Jianyong Hu from the State Key Laboratory of Quantum Optics Technologies and Devices, Institute of Laser Spectroscopy, Shanxi University, together with their collaborators, have proposed a passive quantum compressed sensing (QCS)-based single-photon dynamic imaging technique. This technique leverages the randomness inherent in photon radiation and detection to construct a sensing matrix, thereby enabling frequency-domain sparse signal reconstruction based on discrete and random photon detection events. The detection bandwidth reaches up to 2.05 GHz, representing an improvement of six orders of magnitude compared to traditional photon-counting imaging methods. Furthermore, this method exhibits remarkable robustness against noise through the extraction of dynamic features in the frequency domain of the objects under investigation. As a demonstration of its practical applicability, the research group successfully detected the dynamic frequency characteristics of a drone's rotor at a distance of 10 km, thereby validating the significant application value and technical potential of this technique in real-world scenarios.
Article Title
10-km passive drone detection using broadband quantum compressed sensing imaging