mirror of
https://github.com/c-sooyoung/fold_slice.git
synced 2026-09-18 01:59:11 +09:00
650 lines
28 KiB
Matlab
650 lines
28 KiB
Matlab
% TOMO_SOLVER_DISTRIBUTED tomography solver that loads reconstruction from
|
||
% ptychography, process them and use them to update the current tomgoraphy
|
||
% reconstruction. Then create new estimates of ptychography objects
|
||
%
|
||
% volData = tomo_solver_distributed(par, volData, projData_rec, p0, theta_all, scan_ids)
|
||
%
|
||
% Inputs:
|
||
% **par - parameter structure
|
||
% **volData - initial guess of the tomographic volumes, values are linearized, ie projection = exp(sum(volData,1)) = prod(exp(volData),1)
|
||
% **projData_rec - initial guess of the projections for each angle, probe, positions, etc
|
||
% **p0 - basic parameters for the ptychography solver
|
||
% **theta_all - all angles in degrees
|
||
% **scan_ids - scan numbers that correspons to each angle value
|
||
% *returns*
|
||
% ++volData - final reconstruction of the tomographic volume
|
||
|
||
% Academic License Agreement
|
||
%
|
||
% Source Code
|
||
%
|
||
% Introduction
|
||
% • This license agreement sets forth the terms and conditions under which the PAUL SCHERRER INSTITUT (PSI), CH-5232 Villigen-PSI, Switzerland (hereafter "LICENSOR")
|
||
% will grant you (hereafter "LICENSEE") a royalty-free, non-exclusive license for academic, non-commercial purposes only (hereafter "LICENSE") to use the PtychoShelves
|
||
% computer software program and associated documentation furnished hereunder (hereafter "PROGRAM").
|
||
%
|
||
% Terms and Conditions of the LICENSE
|
||
% 1. LICENSOR grants to LICENSEE a royalty-free, non-exclusive license to use the PROGRAM for academic, non-commercial purposes, upon the terms and conditions
|
||
% hereinafter set out and until termination of this license as set forth below.
|
||
% 2. LICENSEE acknowledges that the PROGRAM is a research tool still in the development stage. The PROGRAM is provided without any related services, improvements
|
||
% or warranties from LICENSOR and that the LICENSE is entered into in order to enable others to utilize the PROGRAM in their academic activities. It is the
|
||
% LICENSEE’s responsibility to ensure its proper use and the correctness of the results.”
|
||
% 3. THE PROGRAM IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, FITNESS FOR
|
||
% A PARTICULAR PURPOSE AND NONINFRINGEMENT OF ANY PATENTS, COPYRIGHTS, TRADEMARKS OR OTHER RIGHTS. IN NO EVENT SHALL THE LICENSOR, THE AUTHORS OR THE COPYRIGHT
|
||
% HOLDERS BE LIABLE FOR ANY CLAIM, DIRECT, INDIRECT OR CONSEQUENTIAL DAMAGES OR OTHER LIABILITY ARISING FROM, OUT OF OR IN CONNECTION WITH THE PROGRAM OR THE USE
|
||
% OF THE PROGRAM OR OTHER DEALINGS IN THE PROGRAM.
|
||
% 4. LICENSEE agrees that it will use the PROGRAM and any modifications, improvements, or derivatives of PROGRAM that LICENSEE may create (collectively,
|
||
% "IMPROVEMENTS") solely for academic, non-commercial purposes and that any copy of PROGRAM or derivatives thereof shall be distributed only under the same
|
||
% license as PROGRAM. The terms "academic, non-commercial", as used in this Agreement, mean academic or other scholarly research which (a) is not undertaken for
|
||
% profit, or (b) is not intended to produce works, services, or data for commercial use, or (c) is neither conducted, nor funded, by a person or an entity engaged
|
||
% in the commercial use, application or exploitation of works similar to the PROGRAM.
|
||
% 5. LICENSEE agrees that it shall make the following acknowledgement in any publication resulting from the use of the PROGRAM or any translation of the code into
|
||
% another computing language:
|
||
% "Data processing was carried out using the PtychoShelves package developed by the Science IT and the coherent X-ray scattering (CXS) groups, Paul
|
||
% Scherrer Institut, Switzerland."
|
||
%
|
||
% Additionally, any publication using the package, or any translation of the code into another computing language should cite
|
||
% K. Wakonig, H.-C. Stadler, M. Odstrčil, E.H.R. Tsai, A. Diaz, M. Holler, I. Usov, J. Raabe, A. Menzel, M. Guizar-Sicairos, PtychoShelves, a versatile
|
||
% high-level framework for high-performance analysis of ptychographic data, J. Appl. Cryst. 53(2) (2020). (doi: 10.1107/S1600576720001776)
|
||
% and for difference map:
|
||
% P. Thibault, M. Dierolf, A. Menzel, O. Bunk, C. David, F. Pfeiffer, High-resolution scanning X-ray diffraction microscopy, Science 321, 379–382 (2008).
|
||
% (doi: 10.1126/science.1158573),
|
||
% for maximum likelihood:
|
||
% P. Thibault and M. Guizar-Sicairos, Maximum-likelihood refinement for coherent diffractive imaging, New J. Phys. 14, 063004 (2012).
|
||
% (doi: 10.1088/1367-2630/14/6/063004),
|
||
% for LSQ-ML:
|
||
% M. Odstrčil, A. Menzel, and M. Guizar-Sicairos, Iterative least-squares solver for generalized maximum-likelihood ptychography, Opt. Express 26(3), 3108 (2018).
|
||
% (doi: 10.1364/OE.26.003108),
|
||
% for mixed coherent modes:
|
||
% P. Thibault and A. Menzel, Reconstructing state mixtures from diffraction measurements, Nature 494, 68–71 (2013). (doi: 10.1038/nature11806),
|
||
% and/or for multislice:
|
||
% E. H. R. Tsai, I. Usov, A. Diaz, A. Menzel, and M. Guizar-Sicairos, X-ray ptychography with extended depth of field, Opt. Express 24, 29089–29108 (2016).
|
||
% (doi: 10.1364/OE.24.029089),
|
||
% and/or for OPRP:
|
||
% M. Odstrcil, P. Baksh, S. A. Boden, R. Card, J. E. Chad, J. G. Frey, W. S. Brocklesby, Ptychographic coherent diffractive imaging with orthogonal probe relaxation.
|
||
% Opt. Express 24.8 (8360-8369) 2016. (doi: 10.1364/OE.24.008360).
|
||
% 6. Except for the above-mentioned acknowledgment, LICENSEE shall not use the PROGRAM title or the names or logos of LICENSOR, nor any adaptation thereof, nor the
|
||
% names of any of its employees or laboratories, in any advertising, promotional or sales material without prior written consent obtained from LICENSOR in each case.
|
||
% 7. Ownership of all rights, including copyright in the PROGRAM and in any material associated therewith, shall at all times remain with LICENSOR, and LICENSEE
|
||
% agrees to preserve same. LICENSEE agrees not to use any portion of the PROGRAM or of any IMPROVEMENTS in any machine-readable form outside the PROGRAM, nor to
|
||
% make any copies except for its internal use, without prior written consent of LICENSOR. LICENSEE agrees to place the following copyright notice on any such copies:
|
||
% © All rights reserved. PAUL SCHERRER INSTITUT, Switzerland, Laboratory for Macromolecules and Bioimaging, 2017.
|
||
% 8. The LICENSE shall not be construed to confer any rights upon LICENSEE by implication or otherwise except as specifically set forth herein.
|
||
% 9. DISCLAIMER: LICENSEE shall be aware that Phase Focus Limited of Sheffield, UK has an international portfolio of patents and pending applications which relate
|
||
% to ptychography and that the PROGRAM may be capable of being used in circumstances which may fall within the claims of one or more of the Phase Focus patents,
|
||
% in particular of patent with international application number PCT/GB2005/001464. The LICENSOR explicitly declares not to indemnify the users of the software
|
||
% in case Phase Focus or any other third party will open a legal action against the LICENSEE due to the use of the program.
|
||
% 10. This Agreement shall be governed by the material laws of Switzerland and any dispute arising out of this Agreement or use of the PROGRAM shall be brought before
|
||
% the courts of Zürich, Switzerland.
|
||
|
||
|
||
function volData = tomo_solver_distributed(par, volData, projData_rec, p0, theta_all, scan_ids)
|
||
import utils.*
|
||
|
||
% be sure to clean GPU first
|
||
reset(gpuDevice)
|
||
|
||
p0.queue.path = par.queue_path;
|
||
|
||
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
|
||
%%%%%%%%%% create the queue %%%
|
||
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
|
||
|
||
system(['rm -rf ', par.queue_path]);
|
||
% if exist(par.queue_path, 'dir')
|
||
% rmdir(par.queue_path, 's');
|
||
% end
|
||
Nangles = length(theta_all);
|
||
|
||
try; mkdir([par.queue_path, '/pending/']); end
|
||
try; mkdir([par.queue_path, '/done/']); end
|
||
|
||
system('rm temp/*/prepared_data.mat ');
|
||
|
||
|
||
% minimum p structure needed to create the queue
|
||
p_init.prepare_data_path = par.prepare_data_path;
|
||
|
||
% create file queue
|
||
for ii = 1:Nangles
|
||
utils.progressbar(ii,length(scan_ids));
|
||
p = p_init;
|
||
p.scan_number = scan_ids(ii);
|
||
save('-v6',fullfile(par.queue_path, sprintf('pending/scan%05d.mat',scan_ids(ii))), 'p');
|
||
end
|
||
|
||
|
||
initial_settings = 'p_initial.mat';
|
||
|
||
save(initial_settings, 'p0');
|
||
|
||
|
||
|
||
Npx_vol = size(volData);
|
||
assert(Npx_vol(1)==Npx_vol(2), 'Horizontal volume size has to be symmetric')
|
||
assert(mod(Npx_vol(1),64) == 0, 'Input volum esize should be dividable by 64')
|
||
thickness = p0.thickness; % Total thickness of the simulated sample if multiple layers are provided;
|
||
|
||
Nangles = length(projData_rec);
|
||
|
||
% choosed optimally distributed the angles in the 180 deg range to minimize overlap
|
||
angle_step = median(diff(theta_all));
|
||
init_angle = math.argmin(abs(theta_all - 180-angle_step*par.downsample_angles/2));
|
||
angle_indices = [1:par.downsample_angles:init_angle-1, init_angle:par.downsample_angles:Nangles];
|
||
volData_hist = {[],[]};
|
||
|
||
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
|
||
|
||
|
||
% initialize one dataset
|
||
[p0] = core.initialize_ptycho(p0);
|
||
p0.prepare_data_path = par.prepare_data_path;
|
||
|
||
|
||
gpu = gpuDevice();
|
||
|
||
|
||
|
||
%% create some loost support mask r
|
||
if par.apply_support
|
||
m_volData = abs(mean(volData,3));
|
||
support_mask = utils.imgaussfilt2_fft(m_volData, 20) > mean(m_volData(:))/2;
|
||
support_mask = utils.imgaussfilt3_conv(support_mask, [10,10,0]);
|
||
volData = volData .* support_mask;
|
||
else
|
||
support_mask = 1;
|
||
end
|
||
%% move arrays to GPU
|
||
par.support_mask = Garray(support_mask);
|
||
par.norm_full = norm(volData(:));
|
||
|
||
% move to uint8 - the volume has to be on GPU !! moving from and to GPU
|
||
% causes too much overhead
|
||
volData = ptychotomo.compress_volume(volData, par);
|
||
volData = Garray(volData);
|
||
|
||
% prepate support masks for the projections using the "support_mask" calculated from the
|
||
% reconstructed tomography volume
|
||
par.weight = sum(support_mask);
|
||
par.weight = utils.crop_pad(par.weight, [1,p0.object_size(2)]);
|
||
par.weight = 1-par.weight / max(par.weight);
|
||
par.support_mask = support_mask;
|
||
|
||
% store all volumes per iteration
|
||
volData_all = cell(par.Niter,1);
|
||
|
||
|
||
|
||
disp('Reconstructing')
|
||
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
|
||
%% iterativelly solve ptycho tomo task %%
|
||
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
|
||
|
||
update_norm = nan(par.Niter, Nangles, par.Niter_inner);
|
||
fourier_error = nan(par.Niter, Nangles,2);
|
||
update_norm_acc = nan(par.Niter, 1);
|
||
hist_numbers = zeros(Nangles,1);
|
||
ptycho_results = cell(Nangles,1); % store temporally data used for reconstruction , avoid data storing
|
||
|
||
|
||
Npx_proj = [p0.object_size, 1];
|
||
|
||
proj_number_list = []; % list of projections yet in procesing
|
||
|
||
|
||
t_plot = tic;
|
||
for iter = 1:par.Niter
|
||
if iter >= max(par.ptycho_reconstruct_start, par.ptycho_ML_reconstruct_start)
|
||
Nlayers = min(2^ceil((iter+1 - par.ptycho_ML_reconstruct_start)), par.Nlayers_max );
|
||
else
|
||
Nlayers = 1;
|
||
end
|
||
|
||
Npx_proj(3) = Nlayers;
|
||
|
||
|
||
full_reconstruction = mod(iter,par.ptycho_interval) == 0 && iter >= par.ptycho_reconstruct_start;
|
||
|
||
fprintf('Free GPU mem: %3.4gGB \n', gpu.AvailableMemory / 1e9)
|
||
|
||
fprintf('Iteration %i/%i Nlayers %i full_reconstruction %i \n', iter, par.Niter, Nlayers, full_reconstruction)
|
||
layer_distance = thickness / Nlayers * ones(1,Nlayers-1) ;
|
||
|
||
Ngroups = length(angle_indices);
|
||
|
||
group_ind = get_golden_ratio_groups(theta_all(angle_indices), Ngroups);
|
||
|
||
|
||
%%%%% start accelerated gradients %%%%%%%%%%%%%%
|
||
if iter == par.ptycho_accel_start
|
||
volData_hist{1} = ptychotomo.decompress_volume(gather(volData), par);
|
||
volData_hist{2} = volData_hist{1};
|
||
elseif iter > par.ptycho_accel_start
|
||
beta = (iter-par.ptycho_accel_start+1)/(iter-par.ptycho_accel_start+3);
|
||
|
||
volData_hist{1} = volData_hist{2};
|
||
volData_hist{2} = ptychotomo.decompress_volume(gather(volData), par);
|
||
|
||
upd = beta*(volData_hist{1}- volData_hist{2});
|
||
update_norm_acc(iter) = norm(upd(:));
|
||
|
||
if ~isfinite(update_norm_acc(iter))
|
||
keyboard
|
||
end
|
||
if all(isfinite(update_norm_acc([iter, iter-1]))) && (update_norm_acc(iter) > update_norm_acc(iter-1))
|
||
% reset the accelerated gradients
|
||
warning('Reset accelerated gradients')
|
||
volData_hist{1} = ptychotomo.decompress_volume(gather(volData), par);
|
||
volData_hist{2} = volData_hist{1};
|
||
par.ptycho_accel_start = iter;
|
||
else
|
||
% else move in the accelerated direction
|
||
volData = volData_hist{2} + upd;
|
||
end
|
||
|
||
volData = ptychotomo.compress_volume(volData, par);
|
||
volData = utils.Garray(volData); % move it back to GPU after gathering for acceleratin computation
|
||
clear upd
|
||
end
|
||
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
|
||
|
||
|
||
verbose(struct('prefix', {'ptychotomo'}))
|
||
|
||
for outer_loop = 1:par.Niter_inner
|
||
fprintf('Outer loop %i \n', outer_loop)
|
||
|
||
tic
|
||
|
||
group_id = 1;
|
||
while group_id <= Ngroups || ~isempty(proj_number_list)
|
||
if verbose() < 0
|
||
utils.progressbar(group_id, Ngroups)
|
||
end
|
||
clear volData_upd_full
|
||
|
||
|
||
verbose(0,'PREPARING')
|
||
try
|
||
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
|
||
%%%%%%%% prepare new initial guess if needed %%%%%%%%%%%%%%%
|
||
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
|
||
|
||
proj_number_list_update = [];
|
||
while length(proj_number_list) < par.max_queue_length && group_id <= Ngroups
|
||
|
||
ind = angle_indices(group_ind{group_id});
|
||
|
||
% calculate how the projection should look like and store the initial guess to "initial_guess.mat"
|
||
for ll = 1:length(ind)
|
||
proj_number = ind(ll);
|
||
scan_number = scan_ids(proj_number);
|
||
|
||
if ~isempty(dir(sprintf('%s/pending/scan%05i.mat', par.queue_path,scan_number)))
|
||
verbose(0, 'Preparing %i', scan_number)
|
||
|
||
verbose(0);
|
||
[projData_model{proj_number}, projData_rec{proj_number}, ptycho_results{proj_number}] = ...
|
||
ptychotomo.prepare_distributed_data(p0, volData, projData_rec{proj_number},layer_distance, par, ...
|
||
iter == par.ptycho_reconstruct_start, iter >= par.ptycho_reconstruct_start );
|
||
|
||
if iter >= par.ptycho_reconstruct_start
|
||
% file is ready for processing, move it to queue
|
||
io.movefile_fast(sprintf('%s/pending/scan%05i.mat', par.queue_path,scan_number), sprintf('%s/scan%05i.mat', par.queue_path,scan_number));
|
||
else
|
||
% skip ptychography reconstrution, pretent that the file is already finished
|
||
io.movefile_fast(sprintf('%s/pending/scan%05i.mat', par.queue_path,scan_number), sprintf('%s/done/scan%05i.mat', par.queue_path,scan_number));
|
||
end
|
||
hist_numbers(proj_number) = hist_numbers(proj_number) +1;
|
||
proj_number_list_update(end+1) = proj_number;
|
||
proj_number_list(end+1) = proj_number;
|
||
continue
|
||
else
|
||
verbose(0, 'Waiting for scan %i to be returned to /pending/', scan_number)
|
||
end
|
||
end
|
||
group_id = group_id + 1;
|
||
|
||
end
|
||
catch err
|
||
keyboard
|
||
end
|
||
|
||
|
||
clear scan_number proj_number
|
||
|
||
%%%%%%% run the reconstruction using the generated queue files %%%
|
||
|
||
|
||
if iter < par.ptycho_reconstruct_start
|
||
verbose(0,'TOMO RECONSTRUCTING')
|
||
for proj_number = proj_number_list_update
|
||
% save previous object estimation and pretend that they were just calculated
|
||
proj = projData_rec{proj_number};
|
||
pout.object{1} = exp(proj.object_c);
|
||
pout.probes = proj.probe;
|
||
pout.positions = proj.positions;
|
||
pout.asize = p0.asize;
|
||
ferr = nan;
|
||
ptycho_results{proj_number} = pout;
|
||
end
|
||
prepare_data_path = sprintf(p0.prepare_data_path, scan_ids(proj_number));
|
||
else
|
||
verbose(0,'PTYCHOTOMO RECONSTRUCTING')
|
||
end
|
||
|
||
|
||
|
||
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
|
||
|
||
|
||
verbose(struct('prefix', {'ptychotomo'}))
|
||
verbose(0,'GATHERING')
|
||
|
||
%%%%%%%%%%%%%% apply the recent reconstructions into the 3D volume %%%%%%
|
||
% load the stored reconstructions and return corresponding object update
|
||
|
||
|
||
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
|
||
%%% start gathering reconstructions %%
|
||
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
|
||
|
||
volData_upd_full = 0;
|
||
proj_number = proj_number_list(1);
|
||
scan_number = scan_ids(proj_number);
|
||
proj_number_list(1) = [];
|
||
scan_file_path = sprintf('%s/done/scan%05d.mat', par.queue_path, scan_number);
|
||
|
||
|
||
timeout_gathering = 0.1; % [s]
|
||
ii = 0;
|
||
while true
|
||
verbose(0)
|
||
if ii * timeout_gathering > par.wait_time_solver
|
||
% skip projetions that were not delivered in more than
|
||
% "wait_time_solver" time
|
||
warning('Failed to gather %s', scan_file_path)
|
||
break
|
||
end
|
||
if exist(scan_file_path, 'file')
|
||
break
|
||
else
|
||
% wait for solver to prepare the reconstruction
|
||
pause(timeout_gathering)
|
||
if ii == 0; verbose(0,'Waiting for reconstruction %i', scan_number); end
|
||
ii = ii + 1;
|
||
continue
|
||
end
|
||
|
||
end
|
||
|
||
if ii * timeout_gathering > par.wait_time_solver
|
||
continue
|
||
end
|
||
|
||
p0.scan_number = scan_number;
|
||
|
||
|
||
verbose(0)
|
||
verbose(0, 'Gathering %i', p0.scan_number)
|
||
|
||
if iter >= par.start_3D_reconstruction
|
||
par.update_step = par.lambda/length(ind)/outer_loop;
|
||
par.update_step = complex(par.update_step / 5, par.update_step); % much slower convergence for amplitude
|
||
else
|
||
par.update_step = 0;
|
||
end
|
||
|
||
try
|
||
if ii * timeout_gathering < par.wait_time_solver
|
||
% calculate and accumulate update of the 3D volume
|
||
[volData,projData_rec{proj_number}, fourier_error(iter,proj_number,:), update_norm(iter,proj_number,outer_loop)] = ...
|
||
ptychotomo.gather_distributed_reconstructions(volData, projData_model{proj_number}, ptycho_results{proj_number},par);
|
||
end
|
||
catch err
|
||
warning('gather_distributed_reconstructions failed: %s', err.message)
|
||
end
|
||
|
||
|
||
|
||
|
||
|
||
% file is already used, move it back to pending ...
|
||
io.movefile_fast(scan_file_path, sprintf('%s/pending/scan%05i.mat', par.queue_path,p0.scan_number));
|
||
|
||
|
||
% release shared memory
|
||
if isa(ptycho_results{proj_number}, 'shm')
|
||
ptycho_results{proj_number}.protected = false; % make it possible to delete the SHM
|
||
ptycho_results{proj_number}.detach;
|
||
ptycho_results{proj_number}.free() ; % free shared memory
|
||
ptycho_results{proj_number} = [];
|
||
end
|
||
|
||
% save memory
|
||
projData_model{proj_number}.object = gather(projData_model{proj_number}.object);
|
||
projData_model{proj_number}.object_c = gather(projData_model{proj_number}.object_c);
|
||
ptycho_results{proj_number} = [];
|
||
|
||
|
||
if math.norm2(projData_rec{proj_number}.probe) > 2.1
|
||
keyboard
|
||
end
|
||
|
||
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
|
||
%% %%%%%%%%%%% PLOT PROGRESS %%%%%%%%%%%%
|
||
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
|
||
|
||
if par.debug || toc(t_plot) > par.plot_every
|
||
plotting.smart_figure(24)
|
||
subplot(2,2,1)
|
||
plotting.imagesc3D(-imag( ptychotomo.decompress_volume(volData(:,:,ceil(end/2)),par)));
|
||
caxis(gather(math.sp_quantile(-imag(volData(:,:,100)), [5e-3, 1-5e-3], 10)))
|
||
axis off image
|
||
colormap bone
|
||
title('Current reconstruction')
|
||
subplot(2,2,2)
|
||
ind_ok = find(any(~isnan(fourier_error(:,:,1)),2));
|
||
plot(ind_ok,fourier_error(ind_ok,:,1), '-')
|
||
hold all
|
||
plot(ind_ok,fourier_error(ind_ok,:,2), '--')
|
||
plot(ind_ok,nanmean(fourier_error(ind_ok,:,1),2), 'k-', 'Linewidth', 2)
|
||
plot(ind_ok,nanmean(fourier_error(ind_ok,:,2),2), 'k--', 'Linewidth', 2)
|
||
hold off
|
||
set(gca, 'xscale', 'log')
|
||
set(gca, 'yscale', 'log')
|
||
grid on
|
||
axis tight
|
||
title('Fourier error (from ptycho)')
|
||
subplot(2,2,3)
|
||
plot(update_norm(:,:,1))
|
||
hold all
|
||
plot(update_norm_acc, 'k--', 'Linewidth', 2)
|
||
plot(nanmean(update_norm(:,:,1),2), 'k', 'Linewidth', 2)
|
||
plot(nanmean(update_norm(:,:,end),2), 'k:', 'Linewidth', 2)
|
||
hold off
|
||
set(gca, 'xscale', 'log')
|
||
set(gca, 'yscale', 'log')
|
||
grid on
|
||
title('Update difference (in tomo)')
|
||
axis tight
|
||
subplot(2,2,4)
|
||
plotting.imagesc3D(exp(crop_pad(sum(projData_rec{proj_number}.object_c,3), [Npx_vol(3), Npx_vol(1)])))
|
||
axis xy off image
|
||
title('Example of a projection')
|
||
%drawnow
|
||
|
||
plotting.smart_figure(15222)
|
||
subplot(2,2,1)
|
||
plotting.imagesc3D(abs(exp(sum(projData_rec{proj_number}.object_c,3))))
|
||
colorbar
|
||
colormap bone
|
||
axis xy off image
|
||
caxis([0.8, 1.1])
|
||
title('Ptycho - abs')
|
||
subplot(2,2,2)
|
||
plotting.imagesc3D(angle(exp(sum(projData_rec{proj_number}.object_c,3))))
|
||
colorbar
|
||
colormap bone
|
||
axis xy off image
|
||
title('Model - phase')
|
||
|
||
subplot(2,2,3)
|
||
plotting.imagesc3D(abs(exp(sum(projData_model{proj_number}.object_c,3))))
|
||
colorbar
|
||
colormap bone
|
||
axis xy off image
|
||
title('Model - abs')
|
||
caxis([0.8, 1.1])
|
||
subplot(2,2,4)
|
||
plotting.imagesc3D(angle(exp(sum(projData_model{proj_number}.object_c,3))))
|
||
colorbar
|
||
colormap bone
|
||
title('Model - phase')
|
||
|
||
axis xy off image
|
||
|
||
|
||
|
||
if par.Niter_inner > 1
|
||
plotting.smart_figure(13123)
|
||
plot(squeeze(update_norm(max(1,iter-1),:,:))')
|
||
hold all
|
||
plot(nanmedian(squeeze(update_norm(max(1,iter-1),:,:)),1) ,'k','Linewidth', 2)
|
||
hold off
|
||
set(gca, 'xscale', 'log')
|
||
set(gca, 'yscale', 'log')
|
||
title(sprintf('Inner loop convergence iter = %i Nlayers = %i', iter, Nlayers))
|
||
axis tight
|
||
end
|
||
|
||
|
||
drawnow
|
||
|
||
t_plot = tic;
|
||
end
|
||
|
||
|
||
end
|
||
|
||
toc
|
||
|
||
|
||
|
||
|
||
end
|
||
|
||
try
|
||
|
||
|
||
volData = ptychotomo.decompress_volume(volData, par);
|
||
|
||
% plotting.smart_figure(123123)
|
||
% subplot(1,2,1)
|
||
% plotting.imagesc3D(log(mean(abs(math.fftshift_2D(fft2(exp(gather(volData))))),3)))
|
||
% axis image off
|
||
% colormap(plotting.colormaps.franzmap)
|
||
% title('FFT of reconstruction')
|
||
% subplot(1,2,2)
|
||
% plotting.imagesc3D(log(mean(abs(math.fftshift_2D(fft2(exp(gather(volData_upd_full))))),3)))
|
||
% axis image off
|
||
% colormap(plotting.colormaps.franzmap)
|
||
% title('FFT of last update')
|
||
% suptitle(sprintf('Iter %i Nlayers %i fullrecons %i', iter, Nlayers, full_reconstruction))
|
||
|
||
|
||
%%
|
||
plotting.smart_figure(234)
|
||
frame = ceil(size(volData,3)/2);
|
||
subplot(1,2,1)
|
||
plotting.imagesc3D(-imag(volData(:,:,frame)));
|
||
caxis(gather(math.sp_quantile(-imag(volData(:,:,frame)), [5e-3, 1-5e-3], 10)))
|
||
axis off image
|
||
colorbar
|
||
colormap bone
|
||
title('Phase')
|
||
subplot(1,2,2)
|
||
plotting.imagesc3D(-real(volData(:,:,frame)));
|
||
caxis(gather(math.sp_quantile(-real(volData(:,:,frame)), [5e-3, 1-5e-3], 10)))
|
||
colorbar
|
||
axis off image
|
||
colormap bone
|
||
title('Absorbtion')
|
||
plotting.suptitle(sprintf('Reconstruction in %i-th iteration', iter))
|
||
%%
|
||
|
||
img = -gather(imag(volData(:,:,frame)));
|
||
img = img / math.sp_quantile(img, 1-5e-3, 1);
|
||
img = max(0, min(1, img));
|
||
if ~exist('./results_ptychotomo', 'dir'); mkdir('results_ptychotomo'); end
|
||
imwrite(gray2ind(img), bone, sprintf('results_ptychotomo/img_3d_ptycho_iter%i_Nlayers_%i.png', iter, Nlayers))
|
||
|
||
|
||
if ~par.debug
|
||
%% save a preview on disk
|
||
% suptitle(sprintf('Iter %i Nlayers %i fullrecons %i', iter, Nlayers, full_reconstruction))
|
||
% print('-dpng', '-f234', '-r300', sprintf('results_ptychotomo/progress_3d_ptycho_iter%i_Nlayers_%i.png', iter, Nlayers))
|
||
end
|
||
|
||
catch err
|
||
warning('Plotting failed')
|
||
keyboard
|
||
end
|
||
|
||
|
||
|
||
volData_all{iter} = gather(volData(:,:,ceil(end/2)));
|
||
|
||
% save memory
|
||
volData_upd_full = [];
|
||
|
||
volData = ptychotomo.compress_volume(volData, par);
|
||
|
||
end
|
||
|
||
|
||
|
||
%% show final results
|
||
|
||
for iter = 1:par.Niter-1
|
||
try
|
||
volupd(:,:,iter) =volData_all{iter+1} - volData_all{iter};
|
||
volprev(:,:,iter) =volData_all{iter};
|
||
end
|
||
end
|
||
|
||
for iter = 1:par.Niter-2
|
||
try
|
||
a = volupd(:,:,iter);
|
||
b = volupd(:,:,iter+1);
|
||
C(iter) = abs(corr(a(:), b(:)));
|
||
end
|
||
end
|
||
|
||
|
||
figure
|
||
subplot(1,2,1)
|
||
plotting.imagesc3D(-imag(volprev))
|
||
colormap bone
|
||
axis off image
|
||
title('Phase')
|
||
subplot(1,2,2)
|
||
plotting.imagesc3D(-real(volprev))
|
||
colormap bone
|
||
axis off image
|
||
title('Absorbtion')
|
||
plotting.suptitle('Reconstruction quality in each iteration')
|
||
|
||
figure
|
||
plotting.imagesc3D(volupd)
|
||
axis off image
|
||
title('Complex volume update in each iteration')
|
||
|
||
|
||
|
||
|
||
end
|
||
|
||
|
||
|