mg_pick.h File Reference

Header file for single particle picking functions. More...

#include "mg_processing.h"
#include "rwimg.h"
#include "marker.h"

Functions
Bparticle *	particles_from_peaks (Bimage *pcc, long bin, double excl_dist, double part_ori, double &fommin, double fommax=1e30, long maxnum=1000000, double pix_min=2, double pix_max=10)
	Finds the peaks in a cross-correlation map corresponding to particles. More...
Bparticle *	particles_pick_cc (Bimage *p, Bimage *ptemp, Bimage *pmask, double hires, double lores, double fommin, double fommax, double excl_dist, long bin=1)
	Picks particles using cross-correlation. More...
Bparticle *	particles_pick_var (Bimage *p, long average_kernel, long var_kernel, double nsig, double part_ori, double excl_dist, long bin)
	Picks particles in variance map. More...
Bparticle *	particles_pick_var (Bimage *p, long average_kernel, long var_kernel, double cutmin, double cutmax, double part_ori, double excl_dist, long bin)
Bparticle *	particles_pick_cc (Bstring &filename, long img_num, Bimage *ptemp, Bimage *pmask, double hires, double lores, double fommin, double fommax, double excl_dist, long bin=1)
	Picks particles using cross-correlation. More...
Bparticle *	particles_pick_var (Bstring &filename, long img_num, long average_kernel, long var_kernel, double nsig, double part_ori, double excl_dist, long bin)
	Picks particles in variance map. More...
double	project_pick_particles (Bproject *project, Bimage *ptemp, Bimage *pmask, double hires, double lores, double fommin, double fommax, double excl_dist, long bin)
	Picks particles using cross-correlation. More...
double	project_pick_particles (Bproject *project, long average_kernel, long var_kernel, double nsig, double part_ori, double excl_dist, long bin)
	Picks particles in variance map. More...
long	project_pick_particles (Bproject *project, double din, double dout, int avg_kernel, double ainc, int flags, int contrast)
	Locates particles using a fore/background difference measure. More...
long	project_pick_background (Bproject *project, long number, long average_kernel, long var_kernel, double excl_dist)
	Picks background areas not overlapping existing particles. More...
long	project_pick_sym_axis (Bproject *project, Bsymmetry &sym, int sym_axis, double axis_dist)
	Picks subregions in 3D particles on the given symmetry axis. More...
double	project_extract_orient_particles (Bproject *project, Bstring &tempfile, Bsymmetry &sym, double hires, double lores, long bin)
	Picks particles using cross-correlation. More...

Detailed Description

Header file for single particle picking functions.

Author

Bernard Heymann

Date

Created: 20000505

Modified: 20190912

Function Documentation

particles_from_peaks()

Bparticle * particles_from_peaks	(	Bimage *	pcc,
		long	bin,
		double	excl_dist,
		double	part_ori,
		double &	fommin,
		double	fommax,
		long	maxnum,
		double	pix_min,
		double	pix_max
	)

Finds the peaks in a cross-correlation map corresponding to particles.

Parameters

*pcc	peak map (after binning).
bin	binning to speed up calculations.
excl_dist	distance between peaks.
part_ori	particle origin.
fommin	minimum threshold to accept peaks.
fommax	maximum threshold to accept peaks.
maxnum	maximum number of peaks to pick.
pix_min	minimum peak width.
pix_max	maximum peak width.

Returns

Bparticle* list of particles.

The map is searched in increments of the particle radius to identify
peaks above the threshold and within a box the size of the
particle radius. The identified peaks are further examined to eliminate 
ones that are too close to a higher scoring peak. The acceptable distance
between peaks is set to 1.8 times the particle radius.

particles_pick_cc() [1/2]

Bparticle * particles_pick_cc	(	Bimage *	p,
		Bimage *	ptemp,
		Bimage *	pmask,
		double	hires,
		double	lores,
		double	fommin,
		double	fommax,
		double	excl_dist,
		long	bin
	)

Picks particles using cross-correlation.

Parameters

*p	image to pick from.
*ptemp	template image.
*pmask	frequency space mask.
hires	high resolution limit.
lores	low resolution limit.
fommin	minimum FOM cutoff.
fommax	maximum FOM cutoff.
excl_dist	minimum distance between particles.
bin	level of image binning.

Returns

Bparticle* list of particles.

A template is cross-correlated with the input image including
bandpass filtering to target the size of the particle.

particles_pick_cc() [2/2]

Bparticle * particles_pick_cc	(	Bstring &	filename,
		long	img_num,
		Bimage *	ptemp,
		Bimage *	pmask,
		double	hires,
		double	lores,
		double	fommin,
		double	fommax,
		double	excl_dist,
		long	bin
	)

Picks particles using cross-correlation.

Parameters

&filename	image to pick from.
img_num	sub-image number.
*ptemp	template image.
*pmask	frequency space mask.
hires	high resolution limit.
lores	low resolution limit.
fommin	minimum FOM cutoff.
fommax	maximum FOM cutoff.
excl_dist	minimum distance between particles.
bin	level of image binning.

Returns

Bparticle* list of particles.

A template is cross-correlated with the input image including
bandpass filtering to target the size of the particle.
The template must have the correct pixel size.

particles_pick_var() [1/3]

Bparticle * particles_pick_var	(	Bimage *	p,
		long	average_kernel,
		long	var_kernel,
		double	cutmin,
		double	cutmax,
		double	part_ori,
		double	excl_dist,
		long	bin
	)

particles_pick_var() [2/3]

Bparticle * particles_pick_var	(	Bimage *	p,
		long	avg_kernel,
		long	var_kernel,
		double	nsig,
		double	part_ori,
		double	excl_dist,
		long	bin
	)

Picks particles in variance map.

Parameters

*p	image to pick from.
avg_kernel	averaging kernel size.
var_kernel	variance kernel size.
nsig	multiple of sigma above variance average to accept peaks.
part_ori	particle origin.
excl_dist	minimum distance between particles.
bin	level of image binning.

Returns

Bparticle* list of particles.

A copy of the micrograph is filtered with an averaging kernel
and a variance map calculated. The variance map is then used to
find high variance peaks as candidate locations for particles.

particles_pick_var() [3/3]

Bparticle * particles_pick_var	(	Bstring &	filename,
		long	img_num,
		long	avg_kernel,
		long	var_kernel,
		double	nsig,
		double	part_ori,
		double	excl_dist,
		long	bin
	)

Picks particles in variance map.

Parameters

&filename	image to pick from.
img_num	sub-image number.
avg_kernel	averaging kernel size.
var_kernel	variance kernel size.
nsig	multiple of sigma above variance average to accept peaks.
part_ori	particle origin.
excl_dist	minimum distance between particles.
bin	level of image binning.

Returns

Bparticle* list of particles.

A copy of the micrograph is filtered with an averaging kernel
and a variance map calculated. The variance map is then used to
find high variance peaks as candidate locations for particles.

project_extract_orient_particles()

double project_extract_orient_particles	(	Bproject *	project,
		Bstring &	tempfile,
		Bsymmetry &	sym,
		double	hires,
		double	lores,
		long	bin
	)

Picks particles using cross-correlation.

Parameters

*project	project parameter structure.
&tempfile	template image.
sym	point group symmetry.
hires	high resolution limit.
lores	low resolution limit.
bin	level of image binning.

Returns

double minimum threshold used to accept peaks.

Each micrograph is cross-correlated with the template image including
bandpass filtering and frequency space masking.

project_pick_background()

long project_pick_background	(	Bproject *	project,
		long	number,
		long	avg_kernel,
		long	var_kernel,
		double	excl_dist
	)

Picks background areas not overlapping existing particles.

Parameters

*project	project parameter structure.
number	maximum number of background images to pick.
avg_kernel	averaging kernel to smooth the image.
var_kernel	kernel to calculate a local variance image.
excl_dist	exclusion distance between areas.

Returns

long number of background areas.

project_pick_particles() [1/3]

double project_pick_particles	(	Bproject *	project,
		Bimage *	ptemp,
		Bimage *	pmask,
		double	hires,
		double	lores,
		double	fommin,
		double	fommax,
		double	excl_dist,
		long	bin
	)

Picks particles using cross-correlation.

Parameters

*project	project parameter structure.
*ptemp	template image.
*pmask	frequency space mask.
hires	high resolution limit.
lores	low resolution limit.
fommin	minimum FOM cutoff.
fommax	maximum FOM cutoff.
excl_dist	minimum distance between particles.
bin	level of image binning.

Returns

double minimum threshold used to accept peaks.

Each micrograph is cross-correlated with the template image including
bandpass filtering and frequency space masking.

project_pick_particles() [2/3]

long project_pick_particles	(	Bproject *	project,
		double	din,
		double	dout,
		int	avg_kernel,
		double	ainc,
		int	flags,
		int	contrast
	)

Locates particles using a fore/background difference measure.

Parameters

*project	project parameter structure.
din	inner diameter corresponding to particle edge.
dout	outer diameter corresponding to background.
avg_kernel	averaging kernle to smooth difference map.
ainc	angular increment.
flags	flags: bit 1 = filter extremes.
contrast	contrast direction (foreground: white=1, black=0).

Returns

long number of particles.

Two marker sets are generated at the indicated diameters.
At each pixel, the marker sets are used to calculate a difference
between the foreground and background as an estimate of the
presence of a particle.

project_pick_particles() [3/3]

double project_pick_particles	(	Bproject *	project,
		long	avg_kernel,
		long	var_kernel,
		double	nsig,
		double	part_ori,
		double	excl_dist,
		long	bin
	)

Picks particles in variance map.

Parameters

*project	project parameter structure.
avg_kernel	averaging kernel size.
var_kernel	variance kernel size.
nsig	multiple of sigma above variance average to accept peaks.
part_ori	particle origin.
excl_dist	minimum distance between particles.
bin	level of image binning.

Returns

double minimum threshold used to accept peaks.

A copy of the micrograph is filtered with an averaging kernel
and a variance map calculated. The variance map is then used to
find high variance peaks as candidate locations for particles.

project_pick_sym_axis()

long project_pick_sym_axis	(	Bproject *	project,
		Bsymmetry &	sym,
		int	sym_axis,
		double	axis_dist
	)

Picks subregions in 3D particles on the given symmetry axis.

Parameters

*project	parameter structure with all parameters.
sym	point group symmetry.
sym_axis	symmetry axis to pick subregions.
axis_dist	distance along symmetry axis.

Returns

long number of new particles.

The existing particles are replaced by the new particles.