Abstract
A laterally translated thin diffuser can not only extend the effective passband of a microscope, but also encode sample information into apparently random intensity measurements before the data leave the imaging system. Leveraging this dual functionality, we propose a unified ptychographic structured modulation (PSM) framework that enables a conventional microscope to support both robust super-resolution imaging and privacy-preserving optical encryption. In the imaging mode, we formulate a reconstruction model that explicitly accounts for free-space propagation, diffuser modulation, and coherent transfer function filtering, and solve the resulting inverse problem by combining amplitude-consistency data fidelity, complex-domain regularization, proximal updates, and momentum acceleration. In the encryption mode, the same diffuser-scanning measurements are compressed and scrambled to form ciphertexts, while authorized recovery is achieved by integrating the calibrated physical model with a plug-and-play denoising prior. Benefiting from this shared modulation interface, the proposed framework remains stable under lateral-position mismatch, measurement noise, and sparse sampling. Super-resolution imaging experiments show that the proposed reconstruction method suppresses artifacts more effectively than the extended ptychographic iterative engine (ePIE), preserves useful details even when the number of measurements is reduced by approximately 61%, and resolves features beyond the diffraction limit of a 0.10 NA objective. Optical encryption experiments further demonstrate that the generated ciphertexts are visually unintelligible, that high-fidelity recovery is achieved with the correct physical key, and that reconstruction rapidly fails under key mismatch. These results indicate that diffuser-based PSM can serve as a shared physical front end for high-resolution quantitative microscopy and secure transmission of pathological information.