Abstract

When the deterministic aberrations are known in an optical system, traditional de-blurring methods, such as the Wiener filter and aberration compensating filters, are effective. However, when the aberrations are difficult to quantify, as are telescope aberrations or the aberrations in the human eye, more practical methods are needed. One potential method for de-blurring a system with unknown aberrations is the bispectral imaging method. This image recovery technique reconstructs the object's Fourier phase by using a sequence of statistically independent aberrated images of the spatially incoherent object. It has been found to be promising in removing the effects of deterministic aberrations when random aberrations are present or are artificially introduced into the system.¹, ² Through computer simulations, we have found the correct amount of random aberrations to have present in a system containing deterministic aberrations to optimize the image quality of the reconstruction at high light levels. Defocus and several third order aberrations were considered in the isoplanatic case. The performance of this method was characterized by reconstructing a point source and computing its Strehl ratio. These results will be used to incorporate the bispectral imaging method as part of a noninvasive technique to reconstruct high resolution images of the eye's fundus in human subjects. Computer simulated images will be presented.

© 1992 Optical Society of America