Neural network based synthetic aperture ladar imaging through marine atmosphere

Abstract

The atmospheric turbulence can severely affect the azimuth imaging of the Synthetic Aperture Ladar (SAL). Based on a neural network, this study generates atmospheric phase screens complied with marine atmosphere spectrum. The SAL imaging results with different turbulence intensities and laser wavelengths are calculated, and the changing law of the ratio of real aperture length to atmospheric coherence diameter with turbulence intensity under different slant ranges and laser wavelengths are analyzed in this paper. The results suggest that the SAL imaging becomes worse as the turbulence intensity is stronger, and even the target is not able to be distinguished. A longer laser wavelength can be used to alleviate the defocus effect, and the phase error compensation algorithm based on marine phase screens can effectively improve the SAL imaging quality to guarantee the real–time process of echo signals of SAL in the marine atmosphere.