|
This research deals with an approach of holistic optimization for optical systems in one open software environment. Studying the available methods of optical design and optimization packages, we understand that all existing approaches have drawbacks and/or missing elements. Hence, we chose to focus on: (i) develop an open optical design software from scratch, (ii) a multi-objective approach that considers not only the image quality, but also the post image processing, (iii) broad exploration of the holistic design space to find the best possible trade-off solutions . In any optical design software process ow, we first need the lens data from the user. Next we trace the rays along the optical system to calculate the OPD map at the exit pupil. Notably, the efficiency and precision of any optical design software depends on the number of rays traced. Here we improved the performance of the design software by using interpolation schemes. We implemented bilinear, cubic, spline interpolation schemes and compared the results. Based on the OPD analysis we estimate the performance of the optical system by calculating the FFT and Huygens PSF, MTF, spot diagram, OPD diagram and Zernike coefficients. Here we also address the sampling problems of the FFT PSF and study the relationship between the sampling of OPD and PSF. In contrast to commercial design software we disclose and extend the underlying algorithms and show the impact on the results. As a main feature we employ image processing libraries to enhance the image inside the optimization iteration. Hence, we build an end to end optical design software in which we allow tolerances to certain optical imaging aberrations and still retrieve the same processed image performance. The final software is an open source and is available to all and anyone can contribute to improve it in future. In summary post processing of the image is an integral part of the optimization of the optical system and therefore allows for an overall simpler but powerful optical system.
|
