基于视觉和光电二极管的跟瞄方法对比(特邀)OA

[Objective]Existing Free-Space Optical(FSO)communication systems predominantly rely on either cameras or Photodiode(PD).However,there is a scarcity of quantitative modeling and comparative analysis regarding these two mecha-nisms.Specifically,the differences in response speed and tracking accuracy under high-dynamic scenarios remain underexplored.Moreover,there is also a lack of quantitative comparisons based on experimental data.[Methods]To address these issues,this paper first establishes a laser spot sampling model for the camera and a mapping model correlating light intensity distribution with spot position for the photodetector.Subsequently,tracking and pointing algorithms tailored for both the camera and the Quadrant Photodiode(QPD)are developed.Furthermore,a hybrid photodetector architecture,comprising a central QPD surrounded by an array of eight peripheral PD,is designed to overcome the limited Field of View inherent to traditional PD devices.Finally,a hard-ware experimental platform simulating high-dynamic scenarios is constructed to conduct closed-loop tracking experiments based on the signals generated by the proposed algorithms.[Results]Experimental results demonstrate that the camera-based system achieves a capture range of approximately 10°,with a convergence time of about 10 s for an initial deviation of 3°,exhibiting supe-rior wide-area coverage capabilities.In contrast,the hybrid photodetector system captures spots within a 2° range and achieves convergence in approximately 1 s for a 1° initial deviation,highlighting its advantage in high-speed response.[Conclusion]These findings conclusively validate the complementary nature of camera-based and PD-based tracking systems from both theoret-ical and experimental perspectives,underscoring the necessity of integrating visual coarse acquisition with photoelectric fine track-ing in practical engineering.This work provides a rigorous theoretical basis and experimental data for constructing robust air-ground optical links.