Improvements on sampling of point spread function in optical transfer function measurement

Point spread function (PSF), which is the intensity distribution of optical system impulse response and usually acquired by imaging a single pinhole in experiment, can characterize the quality of optical imaging system. Faithful recording of the two-dimensional intensity distribution of PSF is key for accurate measurement of optical transfer function (OTF), however distortions in recorded PSF can be easily caused by a large sampling interval and the electronic noise of the detector. Under a given sampling interval, the position-phase difference between pixels and intensity signals can change the intensity distribution of acquired PSF remarkably, making the computed OTF or MTF (modulation transfer function) error prone. Aiming at problems existing in pinhole based MTF measurement methods, this paper developed a new method with underline physics similar to that of slanted edge method to realize sub-pixel sampling of PSF intensity by using accurate non-integer up-sampling matrix of separate binary pixels and applying random patterns shown on digital micro-mirror device (DMD) as target. Numerical simulations show that improvement on the discrete sampling of point spread function with this method is very helpful to improve anti-noise robustness and the accuracy of optical transfer function measurement.