![]() |
Bsoft 2.1.4
Bernard's software package
|
Functions to assess the resoltion of a tomographic tilt series. More...
Functions | |
Bimage * | mg_tomo_resolution (Bproject *project, int micrograph_id, double hi_res, double sampling_ratio, double scale, Vector3< long > size, double fast_angle, int action, double wiener, double cutoff, Bstring &psfile) |
Estimating the resolution of one micrograph in an aligned tilt series. More... | |
vector< Bplot * > | project_tomo_resolution (Bproject *project, double hi_res, double sampling_ratio, double scale, Vector3< long > size, double fast_angle, int action, double wiener, double cutoff) |
Estimating the resolution of the tomogram from an aligned tilt series. More... | |
Bplot * | project_tomo_particle_resolution (Bproject *project, double hi_res, double sampling_ratio, double fast_angle, double cutoff) |
Estimates the resolution for each particle image in each micrograph. More... | |
long | img_pack_2D_into_central_section (Bimage *p, Bimage *prec, Bimage *prec2, long ft_size, double scale, double hi_res, Matrix3 matr, Matrix3 mat, int inplane) |
Packs a 2D Fourier transform into a 3D reciprocal space volume. More... | |
Bplot * | plot_tilt_resolution (Bproject *project) |
Plots the estimated resolution against the tilt angle. More... | |
Functions to assess the resoltion of a tomographic tilt series.
long img_pack_2D_into_central_section | ( | Bimage * | p, |
Bimage * | prec, | ||
Bimage * | prec2, | ||
long | ft_size, | ||
double | scale, | ||
double | hi_res, | ||
Matrix3 | matr, | ||
Matrix3 | mat, | ||
int | inplane | ||
) |
Packs a 2D Fourier transform into a 3D reciprocal space volume.
*p | 2D Fourier transform. |
*prec | 3D central section. |
*prec2 | 3D central section with in-plane micrograph omitted. |
ft_size | Fourier transform size. |
scale | reconstruction scale. |
hi_res | high resolution limit. |
matr | in plane or reference matrix. |
mat | matrix of image being packed. |
inplane | flag to indicate an in plane image. |
The rotation matrix is used to determine the plane in reciprocal space to which the 2D transform data is added. The map is assumed to be cubic and the 2D transform square. The orientation parameters must be written into the image structure. The real space 2D image must be supplied. This is then padded to more than twice its original size, fourier transformed, and packed into the 3D reciprocal space block.
Bimage * mg_tomo_resolution | ( | Bproject * | project, |
int | micrograph_id, | ||
double | hi_res, | ||
double | sampling_ratio, | ||
double | scale, | ||
Vector3< long > | size, | ||
double | fast_angle, | ||
int | action, | ||
double | wiener, | ||
double | cutoff, | ||
Bstring & | psfile | ||
) |
Estimating the resolution of one micrograph in an aligned tilt series.
*project | image processing parameter structure. |
micrograph_id | micrograph number to use for resolution test. |
hi_res | high resolution limit. |
sampling_ratio | ratio for averaging window. |
scale | reconstruction scale. |
size | reconstruction size. |
fast_angle | angle to select micrographs for reconstruction. |
action | flag to apply CTF to projections. |
wiener | Wiener factor. |
cutoff | FRC cutoff. |
&psfile | postscript output file name. |
The orientation parameters, view vector, angle of rotation and origin, must all be set. Each image is padded to at least two times its size and its Fourier transform packed into 3D reciprocal space. The figure-of-merit calculated for each reciprocal space voxel is: sum(w*re)^2 + sum(w*im)^2 FOM = --------------------------- sum(w)*sum(w*(re^2 + im^2)) where re real part im imaginary part w weight (inverse distance of image pixel to closest grid point) For voxels with only one data pixel contributing to it, FOM(0). An image is used in the reconstruction if its selection flag has been set. If the selection number is less than zero, all particles with selection flags greater than zero are used. If the selection number is zero or above, all particles with the selection flag set to the same number are used.
Plots the estimated resolution against the tilt angle.
*project | project structure. |
The resolution estimates must be encoded in the micrograph FOM's.
Bplot * project_tomo_particle_resolution | ( | Bproject * | project, |
double | hi_res, | ||
double | sampling_ratio, | ||
double | fast_angle, | ||
double | cutoff | ||
) |
Estimates the resolution for each particle image in each micrograph.
*project | micrograph project. |
hi_res | high resolution limit for resolution estimation. |
sampling_ratio | ratio for averaging window. |
fast_angle | angle to select micrographs for reconstruction. |
cutoff | FRC cutoff to use. |
Requires the particles to be defined in all micrographs. The NLOO algorithm is used for each particle.
vector< Bplot * > project_tomo_resolution | ( | Bproject * | project, |
double | hi_res, | ||
double | sampling_ratio, | ||
double | scale, | ||
Vector3< long > | size, | ||
double | fast_angle, | ||
int | action, | ||
double | wiener, | ||
double | cutoff | ||
) |
Estimating the resolution of the tomogram from an aligned tilt series.
*project | image processing parameter structure. |
hi_res | high resolution limit. |
sampling_ratio | ratio for averaging window. |
scale | reconstruction scale. |
size | reconstruction size. |
fast_angle | angle to select micrographs for reconstruction. |
action | flag to apply CTF to projections. |
wiener | Wiener factor. |
cutoff | FRC cutoff. |
The orientation parameters, view vector, angle of rotation and origin, must all be set. Each image is padded to at least two times its size and its Fourier transform packed into 3D reciprocal space. The figure-of-merit calculated for each reciprocal space voxel is: sum(w*re)^2 + sum(w*im)^2 FOM = --------------------------- sum(w)*sum(w*(re^2 + im^2)) where re real part im imaginary part w weight (inverse distance of image pixel to closest grid point) For voxels with only one data pixel contributing to it, FOM(0). An image is used in the reconstruction if its selection flag has been set. If the selection number is less than zero, all particles with selection flags greater than zero are used. If the selection number is zero or above, all particles with the selection flag set to the same number are used.