Donglin Ma
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Research Interests
Donglin Ma has a total of 116 co-authors. Here are the top co-authors by number of shared publications:
Yan Mo
21 shared publicationsHuiru Ji
18 shared publicationsShili Wei
18 shared publications- Hao Tan
School of Optical and Electronic Information, Huazhong University of Science and Technologyhttps://ror.org/00p991c53, Wuhan, Hubei 430074, China
16 shared publications Hao Tan
14 shared publications- Yan Mo
Huazhong University of Science and Technology
13 shared publications
Latest Publications
- Open Access
- 10 November 2026
- Published by Springer Nature in Research Square
Toward a Perfect Lens: Automated Design of GRIN Double-Freeform Systems for Broadband Aberration and Chromaticity Correction
In conventional optical design, the inability of a single lens to simultaneously correct high-order geometric and chromatic aberrations over a broad spectral range constitutes a persistent physical bottleneck limiting the development of ultracompact imaging systems. This fundamentally arises from severely restricted degrees of freedom in traditional lenses, which rely solely on interfacial modulation and the intrinsic dispersion of homogeneous media. Here, we propose a global inverse design framework based on gradient-index (GRIN) dual-freeform optics to overcome single-lens constraints. By introducing a wavelength-inclusive extended phase space, the complex ray-tracing evolution is rigorously transformed into an analytical algebraic reconstruction process. This framework reveals a physical mechanism for chromatic aberration compensation: exploiting the three-dimensional spatial dispersion gradient within the GRIN medium induces an anomalous negative chromatic aberration effect. This globally and precisely cancels the intrinsic positive chromatic dispersion introduced by freeform refractions. As methodological validation, we inversely designed an objective singlet lens with a demanding half field of view of 50 degrees. Quantitative evaluations demonstrate that across the entire visible spectrum, the full-field root-mean-square (RMS) spot radius is strictly controlled within 8 nm. This achieves broadband aberration-free and achromatic imaging well within the diffraction limit. This work provides a rigorous theoretical methodology and an implementable pathway for designing ultracompact imaging optics, substantially reducing the reliance on manual experience and iterative numerical optimization in advanced freeform design.
- 10 November 2026
- Published by Springer Nature in Research Square
Open AccessToward a Perfect Lens: Automated Design of GRIN Double-Freeform Systems for Broadband Aberration and Chromaticity Correction
In conventional optical design, the inability of a single lens to simultaneously correct high-order geometric and chromatic aberrations over a broad spectral range constitutes a persistent physical bottleneck limiting the development of ultracompact imaging systems. This fundamentally arises from severely restricted degrees of freedom in traditional lenses, which rely solely on interfacial modulation and the intrinsic dispersion of homogeneous media. Here, we propose a global inverse design framework based on gradient-index (GRIN) dual-freeform optics to overcome single-lens constraints. By introducing a wavelength-inclusive extended phase space, the complex ray-tracing evolution is rigorously transformed into an analytical algebraic reconstruction process. This framework reveals a physical mechanism for chromatic aberration compensation: exploiting the three-dimensional spatial dispersion gradient within the GRIN medium induces an anomalous negative chromatic aberration effect. This globally and precisely cancels the intrinsic positive chromatic dispersion introduced by freeform refractions. As methodological validation, we inversely designed an objective singlet lens with a demanding half field of view of 50 degrees. Quantitative evaluations demonstrate that across the entire visible spectrum, the full-field root-mean-square (RMS) spot radius is strictly controlled within 8 nm. This achieves broadband aberration-free and achromatic imaging well within the diffraction limit. This work provides a rigorous theoretical methodology and an implementable pathway for designing ultracompact imaging optics, substantially reducing the reliance on manual experience and iterative numerical optimization in advanced freeform design.
- Open Access
- 24 October 2026
- Published by Elsevier in SSRN Electronic Journal
Automated Aspheric Zoom Optical Design via Particle Swarm Optimization and Differentiable Ray Tracing
Conventional zoom optical design typically begins with paraxial analysis to establish an initial configuration, while subsequent refinement and aberration correction rely on the experience of optical designers. This paper proposes an automated framework for zoom optical design that combines particle swarm optimization (PSO) with differentiable optical optimization. The framework first performs a global search in the paraxial domain to obtain an initial configuration that satisfies the target focal-length specifications. Starting from this configuration, differentiable optimization progressively expands the field of view and aperture through a curriculum-learning strategy while optimizing image quality. An adaptive illumination-based fine-tuning strategy is subsequently introduced to automatically adjust the clear apertures subject to prescribed relative-illumination constraints, thereby improving marginal-field image quality. The framework requires the target specifications and group configuration as inputs and automatically generates optimized zoom systems. Two optical systems with different group configurations were successfully designed using the proposed framework. At 0.7 normalized field, the adaptive illumination module improved the mean MTF at the target spatial frequency by 24% and 15% in Cases 1 and 2, respectively. These results demonstrate the effectiveness and potential applicability of the proposed framework for automated zoom optical design.
- 24 October 2026
- Published by Elsevier in SSRN Electronic Journal
Open AccessAutomated Aspheric Zoom Optical Design via Particle Swarm Optimization and Differentiable Ray Tracing
Conventional zoom optical design typically begins with paraxial analysis to establish an initial configuration, while subsequent refinement and aberration correction rely on the experience of optical designers. This paper proposes an automated framework for zoom optical design that combines particle swarm optimization (PSO) with differentiable optical optimization. The framework first performs a global search in the paraxial domain to obtain an initial configuration that satisfies the target focal-length specifications. Starting from this configuration, differentiable optimization progressively expands the field of view and aperture through a curriculum-learning strategy while optimizing image quality. An adaptive illumination-based fine-tuning strategy is subsequently introduced to automatically adjust the clear apertures subject to prescribed relative-illumination constraints, thereby improving marginal-field image quality. The framework requires the target specifications and group configuration as inputs and automatically generates optimized zoom systems. Two optical systems with different group configurations were successfully designed using the proposed framework. At 0.7 normalized field, the adaptive illumination module improved the mean MTF at the target spatial frequency by 24% and 15% in Cases 1 and 2, respectively. These results demonstrate the effectiveness and potential applicability of the proposed framework for automated zoom optical design.
- 28 September 2026
- Published by Optica Publishing Group in Optics Express
- Vol. 34  (20) , 37740
Optical design of a transmissive ultra-widefield non-mydriatic confocal fundus imaging system
A fully transmissive ultra-wide-field non-mydriatic confocal fundus imaging system design is proposed for retinal imaging under pupil diameters smaller than 3 mm. Real-ray tracing and conjugate relationship analyses are employed, providing a theoretical basis for the optical design and optimization of the system. Based on this analysis, a fully transmissive coaxial architecture integrating ultra-wide-field imaging and multispectral compatibility is designed. An internal focusing mechanism is further introduced to compensate for refractive errors through axial translation of a focusing lens without changing the overall system length. The proposed design achieves a simulated retinal field angle of 189.06° with a working distance of 15 mm and supports multispectral imaging at 488 nm, 520 nm, 635 nm, and 785 nm. Optical simulation results indicate near diffraction-limited imaging performance across the full field. The retinal resolution remains better than 15 μm over a refractive compensation range from -15 D to +15 D. To the best of our knowledge, the proposed design achieves the largest single-frame field of view for transmissive confocal fundus imaging systems while maintaining non-mydriatic operation.
- 28 September 2026
- Published by Optica Publishing Group in Optics Express
- Vol. 34  (20) , 37740
Optical design of a transmissive ultra-widefield non-mydriatic confocal fundus imaging system
A fully transmissive ultra-wide-field non-mydriatic confocal fundus imaging system design is proposed for retinal imaging under pupil diameters smaller than 3 mm. Real-ray tracing and conjugate relationship analyses are employed, providing a theoretical basis for the optical design and optimization of the system. Based on this analysis, a fully transmissive coaxial architecture integrating ultra-wide-field imaging and multispectral compatibility is designed. An internal focusing mechanism is further introduced to compensate for refractive errors through axial translation of a focusing lens without changing the overall system length. The proposed design achieves a simulated retinal field angle of 189.06° with a working distance of 15 mm and supports multispectral imaging at 488 nm, 520 nm, 635 nm, and 785 nm. Optical simulation results indicate near diffraction-limited imaging performance across the full field. The retinal resolution remains better than 15 μm over a refractive compensation range from -15 D to +15 D. To the best of our knowledge, the proposed design achieves the largest single-frame field of view for transmissive confocal fundus imaging systems while maintaining non-mydriatic operation.
- 22 September 2026
- Published by Optica Publishing Group in Applied Optics
- Vol. 65  (28) , 9656-9673
Simulation of large-angle scattered light propagation in non-rotationally symmetric optical systems
Large-angle scattered light propagation in non-rotationally symmetric optical systems is difficult to simulate because a wide angular spectrum can impose conflicting requirements on spatial sampling and computational-window size. We propose a sectorized angular domain synthesis method (SASM) to address this problem. The target scattering angular domain is determined by downstream aperture acceptance and divided into overlapping local sectors using partition-of-unity weighting to preserve the prescribed angular power distribution. In each sector, the large deflection carrier is removed, the residual field is propagated as a slowly varying envelope in a moving window, and the carrier is restored after registration on a common image plane grid. Sector fields are coherently synthesized within each random-phase realization, followed by Monte Carlo ensemble averaging. The method is evaluated in a tilted two-mirror system under the first-order scattering approximation. Partition-of-unity consistency tests show that the BSDF-weighted sector allocation is independent of the sector number and the overlap ratio within numerical precision. Comparison with an independent direct full-field calculation gives a relative two-dimensional intensity error of 0.849% and an intensity correlation coefficient of 0.999964. Additional sampling tests demonstrate that sectorization suppresses propagation-induced spatial truncation and enables accurate calculation when a single local angular representation exceeds the Nyquist limit. Monte Carlo convergence further confirms a stable ensemble-averaged power estimate. These results demonstrate the numerical reliability of SASM for large-angle scattered light propagation in non-rotationally symmetric optical systems.
- 22 September 2026
- Published by Optica Publishing Group in Applied Optics
- Vol. 65  (28) , 9656-9673
Simulation of large-angle scattered light propagation in non-rotationally symmetric optical systems
Large-angle scattered light propagation in non-rotationally symmetric optical systems is difficult to simulate because a wide angular spectrum can impose conflicting requirements on spatial sampling and computational-window size. We propose a sectorized angular domain synthesis method (SASM) to address this problem. The target scattering angular domain is determined by downstream aperture acceptance and divided into overlapping local sectors using partition-of-unity weighting to preserve the prescribed angular power distribution. In each sector, the large deflection carrier is removed, the residual field is propagated as a slowly varying envelope in a moving window, and the carrier is restored after registration on a common image plane grid. Sector fields are coherently synthesized within each random-phase realization, followed by Monte Carlo ensemble averaging. The method is evaluated in a tilted two-mirror system under the first-order scattering approximation. Partition-of-unity consistency tests show that the BSDF-weighted sector allocation is independent of the sector number and the overlap ratio within numerical precision. Comparison with an independent direct full-field calculation gives a relative two-dimensional intensity error of 0.849% and an intensity correlation coefficient of 0.999964. Additional sampling tests demonstrate that sectorization suppresses propagation-induced spatial truncation and enables accurate calculation when a single local angular representation exceeds the Nyquist limit. Monte Carlo convergence further confirms a stable ensemble-averaged power estimate. These results demonstrate the numerical reliability of SASM for large-angle scattered light propagation in non-rotationally symmetric optical systems.
- 15 September 2026
- Published by Optica Publishing Group in Photonics Research
- Vol. 14  (10) , 4210
High-fidelity image scanning microscopy via programmable line scanning with self-calibration
Image scanning microscopy (ISM) is widely used in the field of super-resolution bioimaging. Compared with confocal microscopy, ISM employs an array detector for signal acquisition, which significantly improves photon utilization efficiency while reducing the complexity of system alignment. However, conventional ISM still suffers from limitations such as slow imaging speed, anisotropic resolution, and low reconstruction fidelity. To address these issues, this study proposes and develops high-fidelity image scanning microscopy (HF-ISM) via programmable line scanning with self-calibration. To comprehensively validate the superior performance of the proposed system, systematic evaluations are conducted through both simulations and experiments. The results demonstrate that, compared with conventional line-scanning ISM, HF-ISM achieves stable sub-diffraction-limited resolution imaging with less sampling data through multi-line synchronous scanning. With orthogonal scanning, it achieves isotropic resolution. By performing self-calibration on the raw data, it attains higher reconstruction fidelity and more uniform quantitative imaging performance across the entire field of view, effectively overcoming the shortcomings of conventional approaches in imaging speed, resolution uniformity, and data fidelity. The proposed HF-ISM framework provides a practical approach for high-speed and high-fidelity fluorescence imaging and may facilitate future developments of ISM-based super-resolution microscopy.
- 15 September 2026
- Published by Optica Publishing Group in Photonics Research
- Vol. 14  (10) , 4210
High-fidelity image scanning microscopy via programmable line scanning with self-calibration
Image scanning microscopy (ISM) is widely used in the field of super-resolution bioimaging. Compared with confocal microscopy, ISM employs an array detector for signal acquisition, which significantly improves photon utilization efficiency while reducing the complexity of system alignment. However, conventional ISM still suffers from limitations such as slow imaging speed, anisotropic resolution, and low reconstruction fidelity. To address these issues, this study proposes and develops high-fidelity image scanning microscopy (HF-ISM) via programmable line scanning with self-calibration. To comprehensively validate the superior performance of the proposed system, systematic evaluations are conducted through both simulations and experiments. The results demonstrate that, compared with conventional line-scanning ISM, HF-ISM achieves stable sub-diffraction-limited resolution imaging with less sampling data through multi-line synchronous scanning. With orthogonal scanning, it achieves isotropic resolution. By performing self-calibration on the raw data, it attains higher reconstruction fidelity and more uniform quantitative imaging performance across the entire field of view, effectively overcoming the shortcomings of conventional approaches in imaging speed, resolution uniformity, and data fidelity. The proposed HF-ISM framework provides a practical approach for high-speed and high-fidelity fluorescence imaging and may facilitate future developments of ISM-based super-resolution microscopy.
- 20 August 2026
- Published by Optica Publishing Group in Optics Express
- Vol. 34  (17) , 32779
Propagation of partially coherent scattered light in non-rotationally symmetric optical systems using mode decomposition
Partially coherent scattered-light propagation in non-rotationally symmetric optical systems is computationally demanding because the cross-spectral density (CSD) is a four-dimensional quantity and the two transverse directions are generally coupled. A coherent-mode-based framework is developed to avoid direct propagation of the full CSD. The input field is represented by a finite set of two-dimensional coherent modes, and surface scattering is introduced through a total-integrated-scatter-normalized complex-amplitude kernel in the angular domain. The resulting weighted scattered components represent the post-scattering CSD without requiring mutual orthogonality. A Gram-matrix-based re-mode decomposition (re-MDM) is further introduced as an optional, error-controlled low-rank compression procedure. The framework is verified against direct CSD propagation and evaluated in a tilted two-mirror reflective system and a transmissive system containing a rotated and decentered rectangular aperture. The numerical studies examine full-rank equivalence, low-rank accuracy, random-phase sensitivity, random-realization convergence, source-coherence dependence, and a non-Gaussian correlation model. Full-rank re-MDM reproduces the direct weighted-component result within numerical accuracy, whereas low-rank re-MDM provides substantially lower intensity errors than direct truncation at the same retained field number. In the tilted two-mirror system, approximately 50 re-decomposed modes retain 99.5% of the scattered-field modal energy, and at R = 80, the global relative L 2 error is reduced by approximately 43 times. These results demonstrate that the proposed framework provides an accurate and efficient approach for partially coherent scattered-light propagation in non-rotationally symmetric optical systems.
- 20 August 2026
- Published by Optica Publishing Group in Optics Express
- Vol. 34  (17) , 32779
Propagation of partially coherent scattered light in non-rotationally symmetric optical systems using mode decomposition
Partially coherent scattered-light propagation in non-rotationally symmetric optical systems is computationally demanding because the cross-spectral density (CSD) is a four-dimensional quantity and the two transverse directions are generally coupled. A coherent-mode-based framework is developed to avoid direct propagation of the full CSD. The input field is represented by a finite set of two-dimensional coherent modes, and surface scattering is introduced through a total-integrated-scatter-normalized complex-amplitude kernel in the angular domain. The resulting weighted scattered components represent the post-scattering CSD without requiring mutual orthogonality. A Gram-matrix-based re-mode decomposition (re-MDM) is further introduced as an optional, error-controlled low-rank compression procedure. The framework is verified against direct CSD propagation and evaluated in a tilted two-mirror reflective system and a transmissive system containing a rotated and decentered rectangular aperture. The numerical studies examine full-rank equivalence, low-rank accuracy, random-phase sensitivity, random-realization convergence, source-coherence dependence, and a non-Gaussian correlation model. Full-rank re-MDM reproduces the direct weighted-component result within numerical accuracy, whereas low-rank re-MDM provides substantially lower intensity errors than direct truncation at the same retained field number. In the tilted two-mirror system, approximately 50 re-decomposed modes retain 99.5% of the scattered-field modal energy, and at R = 80, the global relative L 2 error is reduced by approximately 43 times. These results demonstrate that the proposed framework provides an accurate and efficient approach for partially coherent scattered-light propagation in non-rotationally symmetric optical systems.
- 23 July 2026
- Published by Optica Publishing Group in Optics Express
- Vol. 34  (15) , 28632
Geometrical-physical optics modeling and inverse design for coherent optical field control with freeform elements
We propose a geometrical–physical optics framework for coherent optical field propagation and inverse design in freeform optical systems. By interpreting geometrical optics as the stationary phase approximation (SPA) of physical-optics propagation operators, the proposed framework establishes an asymptotic connection between wave-optics propagation and ray-based optical modeling. A pointwise Fourier transform formulation is introduced to describe the asymptotic mapping between spatial and spatial-frequency domains under the SPA. Based on this formulation, a geometrical–physical optics phase retrieval algorithm is developed by extending the iterative Fourier transform algorithm into regimes with sufficiently large phase gradients. The method is further applied to the inverse design of freeform optical elements for coherent beam shaping. Numerical examples demonstrate that the proposed framework can effectively approximate coherent optical field propagation and achieve coherent beam shaping through freeform inverse design. The proposed approach provides a computationally efficient framework for coherent freeform optical design and may be useful for coherent beam shaping and wavefront engineering applications.
- 23 July 2026
- Published by Optica Publishing Group in Optics Express
- Vol. 34  (15) , 28632
Geometrical-physical optics modeling and inverse design for coherent optical field control with freeform elements
We propose a geometrical–physical optics framework for coherent optical field propagation and inverse design in freeform optical systems. By interpreting geometrical optics as the stationary phase approximation (SPA) of physical-optics propagation operators, the proposed framework establishes an asymptotic connection between wave-optics propagation and ray-based optical modeling. A pointwise Fourier transform formulation is introduced to describe the asymptotic mapping between spatial and spatial-frequency domains under the SPA. Based on this formulation, a geometrical–physical optics phase retrieval algorithm is developed by extending the iterative Fourier transform algorithm into regimes with sufficiently large phase gradients. The method is further applied to the inverse design of freeform optical elements for coherent beam shaping. Numerical examples demonstrate that the proposed framework can effectively approximate coherent optical field propagation and achieve coherent beam shaping through freeform inverse design. The proposed approach provides a computationally efficient framework for coherent freeform optical design and may be useful for coherent beam shaping and wavefront engineering applications.
- 1 July 2026
- Published by Optica Publishing Group in Optics Express
- Vol. 34  (15) , 27232
Differentiable ABCD Matrix Solver for Gaussian Beamlet Decomposition
Gaussian beamlet decomposition is a powerful framework for modeling coherent light propagation in complex optical systems. A key challenge is accurately evaluating the associated ABCD matrix, which is traditionally computed via finite-difference ray tracing with multiple perturbed rays, resulting in numerical errors and sensitivity to perturbation step size. Here, we present a differentiable ABCD matrix solver that enables direct computation of the ABCD matrix without finite-difference approximations. By formulating ray propagation as a differentiable computational graph, the ABCD matrix is obtained via automatic differentiation. We validate the proposed method in both a high-NA microscopic imaging system and a non-imaging beam-shaping system, demonstrating high accuracy and numerical stability compared to conventional approaches. Furthermore, the fully differentiable framework enables gradient-based backpropagation, enabling seamless integration with modern optimization frameworks and inverse wave-optics design.
- 1 July 2026
- Published by Optica Publishing Group in Optics Express
- Vol. 34  (15) , 27232
Differentiable ABCD Matrix Solver for Gaussian Beamlet Decomposition
Gaussian beamlet decomposition is a powerful framework for modeling coherent light propagation in complex optical systems. A key challenge is accurately evaluating the associated ABCD matrix, which is traditionally computed via finite-difference ray tracing with multiple perturbed rays, resulting in numerical errors and sensitivity to perturbation step size. Here, we present a differentiable ABCD matrix solver that enables direct computation of the ABCD matrix without finite-difference approximations. By formulating ray propagation as a differentiable computational graph, the ABCD matrix is obtained via automatic differentiation. We validate the proposed method in both a high-NA microscopic imaging system and a non-imaging beam-shaping system, demonstrating high accuracy and numerical stability compared to conventional approaches. Furthermore, the fully differentiable framework enables gradient-based backpropagation, enabling seamless integration with modern optimization frameworks and inverse wave-optics design.
- 17 June 2026
- Published by Optica Publishing Group in Optics Express
- Vol. 34  (13) , 23656-23674
Double Zernike polynomial-based desensitization design method for off-axis optical systems
High-performance off-axis optical systems require not only excellent nominal performance but also strong robustness against alignment tolerances. However, in conventional desensitization procedures for off-axis systems, the lack of an efficient full-field wavefront representation and sensitivity evaluation framework limits the design efficiency. In this work, a practical desensitization design method is developed by integrating the double Zernike (DZ) polynomial with nodal aberration sensitivity (NAS). The double Zernike expansion is introduced as the full-field representation of the optical path difference (OPD) to address the lack of field dependence in Zernike polynomials. The DZ coefficient matrix is determined through discrete orthogonality and Gaussian quadrature (GQ) with specific sampling conditions in both the field and pupil domains. Based on the DZ coefficient matrix, the full-field wavefront representation is incorporated into an NAS-guided sensitivity evaluation. The full-field wavefront reconstruction capability of DZ polynomials is validated in a TMA (three-mirror anastigmat) system. The DZ-based aberration model is then incorporated into a sensitivity-guided desensitization optimization of an off-axis four-mirror telescope for gravitational-wave detection. Under the same alignment tolerances, the maximum full-field RMS wavefront error of the optimized design at the 90% cumulative probability level is reduced from 0.0843λ to 0.0324λ (λ = 1064 nm). These results demonstrate the effectiveness of the proposed method for full-field wavefront reconstruction and robust design of complex off-axis optical systems.
- 17 June 2026
- Published by Optica Publishing Group in Optics Express
- Vol. 34  (13) , 23656-23674
Double Zernike polynomial-based desensitization design method for off-axis optical systems
High-performance off-axis optical systems require not only excellent nominal performance but also strong robustness against alignment tolerances. However, in conventional desensitization procedures for off-axis systems, the lack of an efficient full-field wavefront representation and sensitivity evaluation framework limits the design efficiency. In this work, a practical desensitization design method is developed by integrating the double Zernike (DZ) polynomial with nodal aberration sensitivity (NAS). The double Zernike expansion is introduced as the full-field representation of the optical path difference (OPD) to address the lack of field dependence in Zernike polynomials. The DZ coefficient matrix is determined through discrete orthogonality and Gaussian quadrature (GQ) with specific sampling conditions in both the field and pupil domains. Based on the DZ coefficient matrix, the full-field wavefront representation is incorporated into an NAS-guided sensitivity evaluation. The full-field wavefront reconstruction capability of DZ polynomials is validated in a TMA (three-mirror anastigmat) system. The DZ-based aberration model is then incorporated into a sensitivity-guided desensitization optimization of an off-axis four-mirror telescope for gravitational-wave detection. Under the same alignment tolerances, the maximum full-field RMS wavefront error of the optimized design at the 90% cumulative probability level is reduced from 0.0843λ to 0.0324λ (λ = 1064 nm). These results demonstrate the effectiveness of the proposed method for full-field wavefront reconstruction and robust design of complex off-axis optical systems.
- 15 June 2026
- Published by Elsevier in Optics and Lasers in Engineering
- Vol. 205
Freeform lens design for illuminance pattern on a large-angle tilted plane
- 15 June 2026
- Published by Elsevier in Optics and Lasers in Engineering
- Vol. 205
Freeform lens design for illuminance pattern on a large-angle tilted plane