Struct Epry 

pub struct Epry {

    pub iterations: usize,
    pub object_step: f64,
    pub pupil_step: f64,
    pub batch_size: usize,
    pub recover_pupil: bool,
    pub constrain_pupil_support: bool,
    pub recover_frame_gains: bool,
    pub gain_step: f64,
    pub minimum_gain: f64,
    pub maximum_gain: f64,
    pub recover_background: bool,
    pub background_step: f64,
    pub minimum_background: f64,
    pub maximum_background: f64,
    pub epsilon: f64,
    pub loss_type: LossType,
}

Embedded pupil-recovery reconstruction for Fourier ptychographic microscopy.

Method

EPRY alternates detector-amplitude projection with two normalized updates: the exit-wave error is divided by pupil power to update the overlapping object-spectrum patch and by object-patch power to update the pupil. Jointly recovering these two complex functions lets the pupil absorb aberrations that would otherwise be imprinted on the reconstructed object. The pupil can be projected back onto the known aperture support after every update.

This implementation also optionally estimates a relative gain and an additive, spatially uniform background for each frame. Those calibration updates and incoherent multiplexing support are crate extensions to the reference EPRY method.

Reference

X. Ou, G. Zheng, and C. Yang, “Embedded pupil function recovery for Fourier ptychographic microscopy,” Optics Express 22(5), 4960–4972 (2014), doi:10.1364/OE.22.004960.

Fields

iterations: usize

Number of complete passes through the acquisition schedule.

object_step: f64

Relaxation factor applied to each object-spectrum correction.

pupil_step: f64

Relaxation factor applied to each pupil correction.

batch_size: usize

Number of measured frames supplied to each reconstruction step.

recover_pupil: bool

Whether to update the complex pupil alongside the object.

constrain_pupil_support: bool

Whether to zero recovered pupil values outside the compiled aperture.

recover_frame_gains: bool

Whether to estimate one multiplicative intensity gain per frame.

gain_step: f64

Fraction of each least-squares frame-gain estimate applied per update.

minimum_gain: f64

Lower bound for recovered frame gains; must be positive.

maximum_gain: f64

Upper bound for recovered frame gains.

recover_background: bool

Whether to estimate one additive, spatially uniform background per frame.

background_step: f64

Fraction of the mean frame residual added to the background per update.

minimum_background: f64

Lower bound for recovered background intensities.

maximum_background: f64

Upper bound for recovered background intensities.

epsilon: f64

Positive numerical floor used in normalized updates.

loss_type: LossType

Loss used for diagnostics; the projection itself always enforces the measured amplitude.

Implementations

impl Epry

pub fn iterations(self, iterations: usize) -> Self

pub fn object_step(self, step: f64) -> Self

pub fn pupil_step(self, step: f64) -> Self

pub fn recover_pupil(self, recover: bool) -> Self

pub fn constrain_pupil_support(self, constrain: bool) -> Self

pub fn recover_frame_gains(self, recover: bool) -> Self

pub fn gain_step(self, step: f64) -> Self

pub fn gain_bounds(self, minimum: f64, maximum: f64) -> Self

pub fn recover_background(self, recover: bool) -> Self

pub fn background_step(self, step: f64) -> Self

pub fn background_bounds(self, minimum: f64, maximum: f64) -> Self

Trait Implementations

impl Clone for Epry

fn clone(&self) -> Epry

Returns a duplicate of the value.

fn clone_from(&mut self, source: &Self)

Performs copy-assignment from source.

impl Debug for Epry

fn fmt(&self, f: &mut Formatter<'_>) -> Result

Formats the value using the given formatter.

impl Default for Epry

fn default() -> Self

Returns the “default value” for a type.

impl ReconstructionAlgorithm for Epry

fn validate(&self) -> Result<()>

fn step<M: MeasurementRead>( &mut self, problem: &ReconstructionProblem<M>, state: &mut ReconstructionState, batch: &Batch, _iteration: usize, ) -> Result<StepDiagnostics>

fn iterations(&self) -> usize

fn batch_size(&self) -> usize

fn validate_problem<M: MeasurementRead>( &self, _problem: &ReconstructionProblem<M>, ) -> Result<()>

fn initialize<M: MeasurementRead>( &self, problem: &ReconstructionProblem<M>, ) -> Result<ReconstructionState>

fn initialize_with_backend<M: MeasurementRead>( &self, problem: &ReconstructionProblem<M>, backend: Arc<dyn Backend>, ) -> Result<ReconstructionState>

fn run<M: MeasurementRead>( self, problem: &ReconstructionProblem<M>, ) -> Result<ReconstructionResult>

where Self: Sized,

fn run_with_callbacks<M: MeasurementRead>( self, problem: &ReconstructionProblem<M>, callbacks: Vec<Box<dyn Callback>>, ) -> Result<ReconstructionResult>

where Self: Sized,

fn run_from_checkpoint<M: MeasurementRead>( self, problem: &ReconstructionProblem<M>, checkpoint: ReconstructionCheckpoint, ) -> Result<ReconstructionResult>

where Self: Sized,