Description of mk

DESCRIPTION

 MK_IMAGE: create eidors image object
   img= mk_image(mdl, elem_data, name)

 Utility function to create an eidors_image object:
 Usage 1:
   img = mk_image( inv_model )  -> uses jacobian backgnd for conductivity

 Usage 2: mdl can be a fwd_model or inv_model
   img = mk_image( mdl, 1 ) -> uniform image with conductivity 1
   img = mk_image( mdl, 2*ones(n_elems,1) ) -> uniform with c=2
   img = mk_image( mdl, 2*ones(n_nodes,1) ) -> image with node data
   img = mk_image( mdl, 1, 'This name') -> Specify a 'name' attribute
   img = mk_image( mdl, ones(n_elems,3) );

 Usage 3: create image from previous image, override conductity
  img = mk_image( other_image, 2 ) -> image with c=2

 Usage 4: create image with specific 'params' properties
  img = mk_image( mdl, 3*ones(64,1),'resistivity' ); % resistivity image

CROSS-REFERENCE INFORMATION

eidors_obj EIDORS_OBJ: maintains EIDORS internals

data_mapper DATA_MAPPER maps img.params data to elem or node data

mk_common_model MK_COMMON_MODEL: make common EIT models

mk_image MK_IMAGE: create eidors image object

num_elems NUM_ELEMS: number of elemnts in a (fwd or inv model or image)

num_nodes NUM_NODES: number of elemnts in a (fwd or inv model or image)

supported_params

calc_jacobian_bkgnd CALC_JACOBIAN_BKGND: calculate background image around

unit_test_cmp UNIT_TEST_CMP: compare matrices in eidors output

left_divide [V] = LEFT_DIVIDE(E,I,tol,pp,V);

calc_noise_params params = GREIT_noise_params(imdl, homg_voltage, sig_voltage)

exponential_covar_prior EXPONENTIAL_COVAR_PRIOR image prior with exponential

np_fwd_solve NP_FWD_SOLVE: data= np_fwd_solve( fwd_model, img)

demo_3d_simdata How to make simulation data using EIDORS3D

eidors2d_demo1 EidorsDemo1 Demonstrates the use of 2D EIT Package with linear basis

ex_fwd2d_high_order ensure dev/m_crabb/forward_problem is on the path

ex_fwd3d_high_order Make common model, and make an image

resistor_model DEMO to show really simple application of EIDORS framework

calc_grid_points CALC_GRID_POINTS - image values at points in grid

calc_slices calc_slices (img, levels, clim ) show slices at levels of an

eidors_colourbar EIDORS_COLOURBAR - create an eidors colourbar with scaling to image

img_point_mapper IMG_POINT_MAPPER - image values at points

mdl_slice_mapper MDL_SLICE_MAPPER: map pixels to FEM elements or nodes

mdl_slice_mesher MDL_SLICE_MESHER A slice of a 3D FEM as a 2D FEM

print_convert PRINT_CONVERT: print figures with anti-aliasing and trim them

show_3d_slices show_3d_slices(img, z_cuts, x_cuts, y_cuts, any_cuts)

show_current SHOW_CURRENT: show current or other quantity defined

show_fem_move SHOW_FEM_MOVE Plot EIT finite element model (FEM) and movement

show_pseudosection SHOW_PSEUDOSECTION: show a pseudo-section image of data

show_slices_move SHOW_SLICES_MOVE Shows planar slices of a 3D FEM with movement vectors

eidors_readimg EIDORS readimg - read reconstructed image files from

sigmatome2_filter SIGMATOME2_FILTER: Hardware filter and stim_patterns for Sigmatome II device

crop_model CROP_MODEL: Crop away parts of a fem model

gmsh_write_mesh gmsh_read_mesh(mdl, [data,] outfile)

mat_idx_to_electrode MAT_IDX_TO_ELECTRODE: create electrodes from mat_idx values

ng_mk_cyl_models NG_MAKE_CYL_MODELS: create cylindrical models using netgen

ng_mk_ellip_models NG_MAKE_ELLIP_MODELS: create elliptical models using netgen

ng_mk_gen_models NG_MK_GEN_MODELS: create generic models using netgen

GREIT3D_distribution GREIT3D_distribution: create target distributions for 3D GREIT

analytic_2d_circle V = analytic_2d_circle(J, [s_h, s_i, b, a, angl])

calc_elem_current calc_elem_current: calculate current vector in each FEM element

calc_model_reduction calc_model_reduction: calculate the fields for a reduced model

check_c2f_quality CHECK_C2F_QUALITY Check the coarse2fine mapping between two models

data_mapper DATA_MAPPER maps img.params data to elem or node data

elem_dim ELEM_DIM: dimension of elements in space (are elements in 2D or 3D space)

get_elem_volume GET_ELEM_VOLUME: VOL = get_elem_volume(fwd_model, map_node )

mdl_dim MDL_DIM: dimension of model space (are nodes in 2D or 3D space)

mk_GREIT_model MK_GREIT_MODEL: make EIDORS inverse models using the GREIT approach

mk_geophysics_model imdl = mk_geophysics_model(str, ne, [option])

mk_grid_c2f MK_GRID_C2F - calculate a coarse2fine mapping for grid coarse models.

mk_grid_model MK_GRID_MODEL: Create reconstruction model on pixelated grid

mk_image MK_IMAGE: create eidors image object

mk_library_model MK_LIBRARY_MODEL - FEM models based on library shapes

mk_lung_image MK_LUNG_IMAGE create an image with lung and heart contrasts from a model

mk_pixel_slice MK_PIXEL_SLICE create a pixel model to reconstruct on

mk_tri2tet_c2f MK_TRI2TET_C2F - coarse2fine mapping between tri-based and tet-based models

mk_tri_c2f MK_TRI_C2F - calculate a coarse2fine mapping for triangle-based models.

mk_voxel_volume MK_VOXEL_VOLUME create a voxel model to reconstruct on

mono2stim MONO2SIM transform monopolar stimulation into another stim pattern

num_elecs NUM_ELECS: number of electrodes attached to model

num_elems NUM_ELEMS: number of elemnts in a (fwd or inv model or image)

num_nodes NUM_NODES: number of elemnts in a (fwd or inv model or image)

select_imdl SELECT_IMDL: select pre-packaged inverse model features

simulate_movement SIMULATE_MOVEMENT simulate small conductivity perturbations

solve_RM_2Dslice SOLVE_RM_2DSLICE: cut slices out of a 3D model

stim_meas_list STIM_MEAS_LIST: mk stimulation pattern from list of electrodes

test_GREIT_model

calc_transferimpedance CALC_TRANSFERIMPEDANCE: Calculates transfer impedance matrix

eit_spice function spice = eit_spice(img, [name])

fwd_solve_1st_order FWD_SOLVE_1ST_ORDER: data= fwd_solve_1st_order( img)

fwd_solve_2p5d_1st_order FWD_SOLVE_2P5D_1ST_ORDER: data= fwd_solve_2p5d_1st_order( img)

fwd_solve_apparent_resistivity fwd_solve_apparent_resistivity: fwd_solve output as apparent resistivity

fwd_solve_halfspace FWD_SOLVE_HALFSPACE: data = fwd_solve_halfspace(img)

fwd_solve_higher_order Solve for voltages (nodes/electrodes) for a forward model.

jacobian_adjoint JACOBIAN_ADJOINT: J= jacobian_adjoint( img )

jacobian_adjoint_2p5d_1st_order JACOBIAN_ADJOINT_2P5D: J= jacobian_adjoint_2p5d_1st_order( img )

jacobian_adjoint_higher_order Find the Jacobian associated with an image (and forward model)

jacobian_apparent_resistivity jacobian_apparent_resistivity: Jacobian output as apparent resistivity

jacobian_movement JACOBIAN_MOVEMENT Computes the Jacobian matrix for conductivity and

jacobian_movement_2p5d_1st_order JACOBIAN_MOVEMENT_2P5D: J = jacobian_movement_2p5d_1st_order( img )

system_mat_1st_order SYSTEM_MAT_1ST_ORDER: SS= system_mat_1st_order( fwd_model, img)

system_mat_2p5d_1st_order SYSTEM_MAT_2P5D_1ST_ORDER: 2.5D system matrix

system_mat_fields SYSTEM_MAT_FIELDS: fields (elem to nodes) fraction of system mat

system_mat_higher_order Assemble the total stiffness matrix : s_mat.E=At;

system_mat_instrument SYSTEM_MAT_INSTRUMENT:

fwd_solve FWD_SOLVE: calculate data from a fwd_model object and an image

inv_solve INV_SOLVE: calculate imag from an inv_model and data

GREIT_desired_img_FEMmesh GREIT_DESIRED_IMG_FEMmesh GREIT onto a FEM mesh

GREIT_desired_img_sigmoid GREIT_DESIRED_IMG_SIGMOID sigmoid-decay desired image function for GREIT

calc_TSVD_RM CALC_TSVD_RM: Calculated truncated Jacobian SVD reconstruction matrix

calc_image_SNR % CALC_IMAGE_SNR: Calculates the signal-to-noise ratio (SNR) in the image

calc_noise_figure CALC_NOISE_FIGURE: calculate the noise amplification (NF) of an algorithm

inv_solve_TSVD INV_SOLVE_TSVD: inverse solver based on truncatated SVD

inv_solve_abs_annealingMetropolis_params INV_SOLVE_ABS_ANNEALINGSIMPLEX_PARAMS absolute solver using the simplex annealing method.

inv_solve_abs_annealingSimplex_params INV_SOLVE_ABS_ANNEALINGSIMPLEX_PARAMS absolute solver using the simplex annealing method.

inv_solve_cg function img= inv_solve_cg( inv_model, data1);

inv_solve_conj_grad INV_SOLVE_CONJ_GRAD inverse solver based on the CG

inv_solve_core INV_SOLVE_CORE Solver using a generic iterative algorithm

inv_solve_gn function img= inv_solve_gn( inv_model, data1);

mk_GN_model MK_GN_MODEL: make EIDORS inverse models using the GREIT approach

prior_exponential_covar PRIOR_EXPONENTIAL_COVAR image prior with exponential

test_3d_resistor Create 3D model of a Rectangular resistor

test_performance TEST_PERFORMANCE: test of difference reconstruction algorithms

test_performance_img TEST_PERFORMANCE: test of difference reconstruction algorithms

SOURCE CODE

0001 function img= mk_image(mdl, elem_data, params, name)
0002 % MK_IMAGE: create eidors image object
0003 %   img= mk_image(mdl, elem_data, name)
0004 %
0005 % Utility function to create an eidors_image object:
0006 % Usage 1:
0007 %   img = mk_image( inv_model )  -> uses jacobian backgnd for conductivity
0008 %
0009 % Usage 2: mdl can be a fwd_model or inv_model
0010 %   img = mk_image( mdl, 1 ) -> uniform image with conductivity 1
0011 %   img = mk_image( mdl, 2*ones(n_elems,1) ) -> uniform with c=2
0012 %   img = mk_image( mdl, 2*ones(n_nodes,1) ) -> image with node data
0013 %   img = mk_image( mdl, 1, 'This name') -> Specify a 'name' attribute
0014 %   img = mk_image( mdl, ones(n_elems,3) );
0015 %
0016 % Usage 3: create image from previous image, override conductity
0017 %  img = mk_image( other_image, 2 ) -> image with c=2
0018 %
0019 % Usage 4: create image with specific 'params' properties
0020 %  img = mk_image( mdl, 3*ones(64,1),'resistivity' ); % resistivity image
0021 
0022 % (C) 2008-12 Andy Adler and Bartlomiej Grychtol.
0023 % Licenced under GPL version 2 or 3
0024 % $Id: mk_image.m 6161 2021-11-02 14:44:49Z bgrychtol $
0025 
0026 if ischar(mdl) && strcmp(mdl,'UNIT_TEST'); do_unit_test; return; end
0027 
0028 default_params = no_params; % later: conductivity ?
0029 default_name    = 'Created by mk_image';
0030 if nargin<3
0031     name = default_name;
0032     params = default_params;
0033 elseif nargin < 4 % preserve old interface
0034     if ismember(params,supported_params)
0035         name = default_name;
0036     else
0037         name = params;
0038         params = default_params;
0039     end
0040 end
0041 
0042 % if nargin<2;
0043 %    try
0044 %      elem_data = mdl.jacobian_bkgnd.value;
0045 %    catch
0046 %      error('mk_image: for one parameter, needs a mdl.jacobian_bkgnd field');
0047 %    end
0048 % end
0049 
0050 switch mdl.type
0051    case 'inv_model'
0052       if nargin == 1
0053          img = calc_jacobian_bkgnd(mdl);
0054          return
0055       else
0056          mdl = mdl.fwd_model;
0057          % warning('Ignoring inv_model.jacobian_bkgnd in favour of the supplied elem_data');
0058       end
0059    case 'fwd_model'
0060       mdl = mdl; % keep model
0061    case 'image'
0062       if nargin == 1
0063          img = data_mapper(mdl);
0064          return
0065       end
0066       mdl = mdl.fwd_model;
0067    otherwise; error('mk_image: no inv_model, fwd_model or image object');
0068 end
0069 
0070 img = eidors_obj('image',name);
0071 img.fwd_model = mdl;
0072 if isfield(mdl,'show_slices');
0073     img.show_slices = mdl.show_slices;
0074 end
0075 img = fill_in_data(img,elem_data,params);
0076 % img.current_params = params;
0077 
0078 function str = no_params
0079 str = 'unspecified';
0080 
0081 function img = fill_in_data(img,elem_data,params)
0082   % If image is presented as row vector, then transpose
0083   %  Shouldn't happen, but for matlab bugs
0084   if size(elem_data,1) == 1 && ...
0085      size(elem_data,2) >  1 
0086      elem_data = elem_data.';
0087   end
0088   switch size(elem_data,1)
0089      case 1
0090         sz = size(img.fwd_model.elems,1);
0091         elem_data_mat =    NaN*ones(sz,1);
0092         elem_data_mat(:) = elem_data;
0093         if strcmp(params,no_params);
0094             img.elem_data = elem_data_mat;
0095         else
0096             img.(params).elem_data = elem_data_mat;
0097         end
0098 
0099      case num_elems( img )
0100         if strcmp(params,no_params);
0101             img.elem_data = elem_data;
0102         else
0103             img.(params).elem_data = elem_data;
0104         end
0105           
0106       case num_nodes( img )
0107         if strcmp(params,no_params);
0108             img.node_data = elem_data;
0109         else
0110             img.(params).node_data = elem_data;
0111         end
0112         
0113       case size(img.fwd_model.coarse2fine,2)
0114         if strcmp(params,no_params);
0115             img.elem_data = elem_data;
0116         else
0117             img.(params).elem_data = elem_data;
0118         end
0119 
0120      otherwise
0121         error('Don''t understand number of elements.');
0122   end
0123 
0124 % TESTS:
0125 function do_unit_test
0126    imdl = mk_common_model('a2c2',8);
0127    im0 = mk_image( imdl );
0128    unit_test_cmp('im1',im0.elem_data, ones(64,1) );
0129 
0130    im0 = mk_image( imdl, 2 ); % warning
0131    unit_test_cmp('im2',im0.elem_data, 2*ones(64,1) );
0132 
0133    im0 = mk_image( imdl.fwd_model, 3*ones(64,1) );
0134    unit_test_cmp('im3',im0.elem_data, 3*ones(64,1) );
0135 
0136    im0 = mk_image(im0); % no change
0137    unit_test_cmp('im4',im0.elem_data, 3*ones(64,1) );
0138 
0139    im0.conductivity.elem_data = im0.elem_data;
0140    im0 = rmfield(im0, 'elem_data');
0141    im1 = mk_image(im0); % use data_mapper
0142    unit_test_cmp('im5',im1.elem_data, 3*ones(64,1) );
0143 
0144    im0.resistivity.node_data = 5;
0145    im0.data_mapper = 'resistivity';
0146    im1 = mk_image(im0); % use data_mapper
0147    unit_test_cmp('im6',im1.node_data(5), 5 );
0148 
0149    imdl.jacobian_bkgnd = rmfield(imdl.jacobian_bkgnd, 'value');
0150    imdl.jacobian_bkgnd.conductivity = im0.conductivity;
0151    im2 = mk_image(imdl);
0152    unit_test_cmp('im7',im2.conductivity.elem_data, 3*ones(64,1) );
0153 
0154    im0 = mk_image( imdl.fwd_model, 3*ones(64,1),'my_name' );
0155    unit_test_cmp('im8a',im0.elem_data, 3*ones(64,1) );
0156 
0157    im0 = mk_image( imdl.fwd_model, 3*ones(64,1),'resistivity' );
0158    unit_test_cmp('im8b',im0.resistivity.elem_data, 3*ones(64,1) );
0159 
0160    im0 = mk_image( imdl.fwd_model, 2*ones(64,3),'resistivity' );
0161    unit_test_cmp('im8c',im0.resistivity.elem_data, 2*ones(64,3) );
0162 
0163    im0 = mk_image( imdl.fwd_model, 3*ones(64,1),'resistivity','my name' );
0164    unit_test_cmp('im9a',im0.resistivity.elem_data, 3*ones(64,1) );
0165    unit_test_cmp('im9b',im0.name, 'my name' );

mk_image

PURPOSE

SYNOPSIS

DESCRIPTION

CROSS-REFERENCE INFORMATION

SUBFUNCTIONS

SOURCE CODE