The paper introduces an Adaptive Cascade Calibrated (ACC) multi-plane phase retrieval technique that addresses the challenge of misalignment in experimental setups. Unlike existing methods that rely on precise alignment, the ACC method implements a computational self-calibration during the phase reconstruction process.
The key components of the ACC method are:
Autofocusing: Accurately determines the locations of each set of measurements and acquires a refocused image of the target sample for calibration.
Adaptive Cascade Calibration: Identifies feature points between adjacent measurements using the Scale Invariant Feature Transform (SIFT) algorithm and computes the homography matrix to calculate the affine matrix between the first and nth measurement. This enables digital calibration of the measurements.
Multi-Plane Phase Retrieval: Conducts an energy-conserved Gerchberg-Saxton (GS) algorithm to perform the final phase retrieval, propagating back and forth between the measurements.
The proposed ACC method eliminates the need for markers or meticulous alignment in experimental setups, preserving the simplicity and cost-effectiveness of multi-plane phase retrieval. It also performs the calibration in the object space rather than the measurement space, reducing the impact of diffraction effects. The effectiveness of the ACC method is validated through simulations and real-world optical experiments involving biological samples, demonstrating its ability to achieve high-quality reconstructions even in the presence of significant misalignment.
다른 언어로
소스 콘텐츠 기반
arxiv.org
더 깊은 질문