Development of novel technologies for High Space-Bandwidth Product Imaging Using Conventional Off-the-Shelf Lenses

Abstract

The ever-increasing capabilities of high-resolution imaging sensors have exposed a significant limitation in conventional optical systems: the inherent constraints of objective lenses that prevent full utilization of these sensors for large field-of-view (FOV) imaging. Objective lenses, while critical for magnification and resolution, cannot realize their full potential in wide-area imaging due to optical aberrations, particularly field curvature. This aberration causes images of flat objects to appear out of focus at the periphery when projected onto a flat detector, compromising image quality across the expanded FOV. This thesis presents two innovative and practical approaches to overcome these limitations and enhance the space-bandwidth product (SBP) of imaging systems using standard objective lenses: a deconvolution-based correction method and an adaptive optics (AO)-based correction method. Both strategies aim to mitigate the effects of optical aberrations without the need for expensive custom- designed lenses or complex optical components. In the first approach, we reconfigure a standard infinity-corrected microscope by adjusting the position of the tube lens relative to the objective lens, creating a short-infinity space configuration. This modification prevents the cutoff of peripheral rays, effectively expanding the usable FOV from the conventional 1.2 mm to approximately 7 mm in diameter without altering the system's magnification or resolution. To correct the aberrations introduced by field curvature over this expanded FOV, we implement a spatially variant deconvolution algorithm. By measuring the point spread functions across the imaging area, we significantly enhance image quality, achieving near-diffraction-limited resolution in central regions and restoring sharpness in peripheral areas. This method demonstrates that computational techniques can effectively compensate for optical imperfections, maximizing the potential of high-resolution sensors with existing optical hardware. The second approach utilizes a sensorless adaptive optics system incorporating a deformable mirror to correct aberrations in real time. We develop a wide-field imaging setup using a 4f configuration with a low magnification, large aperture objective lens and integrate a galvanometer scanner to tile the imaging area, effectively covering an 8.1 mm diameter FOV. By focusing on correcting primary aberrations through iterative optimization, the AO system enhances image quality uniformly across the entire FOV. Bright-field imaging experiments reveal significant improvements, particularly in peripheral regions where aberrations are most pronounced. This approach highlights the effectiveness of real-time wavefront correction in addressing spatially varying aberrations, offering immediate enhancement of raw image quality without extensive post-processing. Our results demonstrate that both methods effectively expand the usable FOV and enhance image quality in imaging systems employing standard objective lenses. The deconvolution-based method is advantageous for applications where post-processing is acceptable and computational resources are available, offering significant improvements using existing equipment. The adaptive optics-based method is ideal for real-time imaging needs, providing immediate correction of aberrations and enhancing image uniformity across large FOVs, albeit requiring specialized hardware. These findings have significant implications for the development and optimization of optical imaging systems. By addressing the inherent limitations of objective lenses and overcoming optical aberrations like field curvature, we enable the capture of more detailed and expansive images without the necessity for costly custom lenses. This advancement is crucial for various scientific and industrial applications that rely on high-quality imaging over large areas. In conclusion, this thesis contributes to the field of optical imaging by presenting practical, cost- effective solutions to transcend the fundamental limitations of objective lenses in wide-FOV imaging. By employing computational deconvolution and adaptive optics correction methods, we demonstrate that it is possible to significantly enhance the space-bandwidth product of imaging systems using standard components. This work paves the way for more accessible high-resolution, wide-area imaging, facilitating innovation and discovery in disciplines that depend on detailed visual analysis.