Enhancing the target recognition rate of orbital angular momentum spectra under cloud and fog conditions via an intensity-phase correction network

Vortex beams carry orbital angular momentum (OAM), and after reflecting from a target, their OAM spectra encode target-related features, thereby facilitating target recognition and reducing data complexity. However, the transmission of vortex beams is often degraded by cloud and fog, which affect both the intensity and phase of the beam, thereby causing dispersion in the OAM spectrum and reducing the recognition rate. To improve the recognition rate, we proposed an intensity–phase correction network (Int-PhaseCorrectionNet) to compensate for complex amplitude distortions caused by cloud and fog interference, thereby reducing OAM spectrum dispersion. Furthermore, a target recognition network was employed to accurately identify various typical target shapes under cloud and fog conditions. The results demonstrated that the proposed method improved the recognition rate from 29.50% to 99.62%. To further expand the application scope of the proposed method, we extended this to complex character-shaped targets, recognition rate increased from 33.25% to 99.88%. These results indicate that the proposed method can effectively correct the intensity, phase, and OAM spectrum in cloudy and foggy environments, significantly enhancing target recognition performance, with potential applications in laser detection and autonomous driving.