The Application For Sparse Aperture Optical System Based On Multispectral Images Phase Diversity

The sparse aperture optical system is compounded by sub-mirrors according to a certain way, which obtains same definition as a single large aperture system, and makes the whole system lighter, smaller and cheaper. But it is difficulty for the systems to avoid piston errors and tilt errors in the process of mounting, which makes wave-front distortion serious, and then reduces the image quality. Phase diversity is based on image incoherent measurement technology which provides real-time data for the wave-front corrector by detecting the wave-front distortion of the system based on non-coherent detection technology, and makes the imaging quality near to the diffraction limited status.The traditional phase diversity is researched under the monochromatic light, with the development of multispectral imaging technology, more information will be obtained on the side of imaging research, it is meaningful to expand the research of phase diversity from monochromatic light to multispectral images. This paper researches phase diversity based on multispectral images, and detects the wave-front error of annular sparse aperture optical system by using different waveband images.Firstly, this paper introduces two structures of the sparse aperture imaging system,analyzes the fill factor, modulation transfer function and the relationship of them using the annular sparse aperture as an example, simulates imaging with different fill factors. Secondly, it analyzes the influence of relative ambiguity on imaging systems,which is caused by piston errors and tilt errors in the process of mounting. This paper estimates the adjust errors of the optical system by phase diversity with the monochromatic light, and then recovers the image. Finally it expands the research of phase diversity from monochromatic light to multispectrum based on multispectral transfer function theory, estimates the wave-front error of the annular sparse aperturesystem, and then recovers the target image.Results indicate that the error values estimated by phase diversity are all lower than 3%, its validity is checked by simulation with phase diversity of multispectral images. In the end it presents that lower relative error may be achieved under the green wavebands, which provides basis for researching the relationship between accuracy recovered by phase diversity and wavelengths.