Hdrl_correlation

Functions
hdrl_xcorrelation_result *	hdrl_xcorrelation_result_wrap (cpl_array *x_corr, const cpl_size max_idx, const cpl_size half_window)
	Constructor for hdrl_xcorrelation_result.
void	hdrl_xcorrelation_result_delete (hdrl_xcorrelation_result *self)
	Destructor for hdrl_xcorrelation_result.
cpl_size	hdrl_xcorrelation_result_get_peak_pixel (const hdrl_xcorrelation_result *self)
	Get the index where the cross correlation reaches its maximum.
double	hdrl_xcorrelation_result_get_peak_subpixel (const hdrl_xcorrelation_result *self)
	Get the index where the cross correlation reaches its maximum, with sub-pixel precision.
cpl_size	hdrl_xcorrelation_result_get_half_window (const hdrl_xcorrelation_result *self)
	Get the half_window used to calculate the cross-correlation.
double	hdrl_xcorrelation_result_get_sigma (const hdrl_xcorrelation_result *self)
	Get the estimated standard deviation of the correlation.
const cpl_array *	hdrl_xcorrelation_result_get_correlation (const hdrl_xcorrelation_result *self)
	Getter for the cross correlation.
hdrl_xcorrelation_result *	hdrl_compute_xcorrelation (const cpl_array *arr1, const cpl_array *arr2, const cpl_size half_window, const cpl_boolean normalize)
	Calculate cross-correlation.
hdrl_xcorrelation_result *	hdrl_compute_offset_gaussian (const cpl_array *arr1, const cpl_array *arr2, const cpl_size half_win, const cpl_boolean normalize, const double bin, const double wrange)
	Calculate gaussian fit on cross-correlation, does a second fitting for refinement.
hdrl_xcorrelation_result *	hdrl_compute_offset_gaussian_internal (const cpl_array *arr1, const cpl_array *arr2, const cpl_size half_win, const cpl_boolean normalize, const double bin, const double wrange)
	Calculate gaussian fit on cross-correlation.

Detailed Description

Function Documentation

hdrl_compute_offset_gaussian()

hdrl_xcorrelation_result * hdrl_compute_offset_gaussian	(	const cpl_array *	arr1,
		const cpl_array *	arr2,
		const cpl_size	half_win,
		const cpl_boolean	normalize,
		const double	bin,
		const double	wrange
	)

Calculate gaussian fit on cross-correlation, does a second fitting for refinement.

Parameters

arr1	First array
arr2	Second array
half_win	half search window where the correlation is calculated
normalize	CPL_TRUE normalize correlation in mean and rms
bin	wavelength bin
wrange	half window wavelength range where the fit is going to be done

Returns

gaussian fit for cross correlation.

hdrl_compute_offset_gaussian_internal()

hdrl_xcorrelation_result * hdrl_compute_offset_gaussian_internal	(	const cpl_array *	arr1,
		const cpl_array *	arr2,
		const cpl_size	half_win,
		const cpl_boolean	normalize,
		const double	bin,
		const double	wrange
	)

Calculate gaussian fit on cross-correlation.

Parameters

arr1	First array
arr2	Second array
half_win	half search window where the correlation is calculated
normalize	CPL_TRUE normalize correlation in mean and rms
bin	wavelength bin
wrange	half window wavelength range where the fit is going to be done

Returns

gaussian fit for cross correlation.

hdrl_compute_xcorrelation()

hdrl_xcorrelation_result * hdrl_compute_xcorrelation	(	const cpl_array *	arr1,
		const cpl_array *	arr2,
		const cpl_size	half_window,
		const cpl_boolean	normalize
	)

Calculate cross-correlation.

Parameters

arr1	First array
arr2	Second array
half_window	half search window where the correlation is calculated
normalize	CPL_TRUE normalize correlation in mean and rms

Returns

cross correlation and index where the peak is. NULL in case of error.

Note

: elements marked as invalid in arr1 or arr2 will be treated as they were out-of-boudary pixels.

hdrl_xcorrelation_result_delete()

void hdrl_xcorrelation_result_delete

(

hdrl_xcorrelation_result *

self

)

Destructor for hdrl_xcorrelation_result.

Parameters

self	hdrl_xcorrelation_result to be deleted

hdrl_xcorrelation_result_get_correlation()

const cpl_array * hdrl_xcorrelation_result_get_correlation

(

const hdrl_xcorrelation_result *

self

)

Getter for the cross correlation.

Parameters

self	hdrl_xcorrelation_result the getter will extract the data from

Returns

the cross correlation

hdrl_xcorrelation_result_get_half_window()

cpl_size hdrl_xcorrelation_result_get_half_window

(

const hdrl_xcorrelation_result *

self

)

Get the half_window used to calculate the cross-correlation.

Parameters

self	hdrl_xcorrelation_result the getter will extract the data from

Returns

the half_window used to calculate the cross-correlation.

hdrl_xcorrelation_result_get_peak_pixel()

cpl_size hdrl_xcorrelation_result_get_peak_pixel

(

const hdrl_xcorrelation_result *

self

)

Get the index where the cross correlation reaches its maximum.

Parameters

self	hdrl_xcorrelation_result the getter will extract the data from

Returns

the index where the cross correlation reaches its maximum

hdrl_xcorrelation_result_get_peak_subpixel()

double hdrl_xcorrelation_result_get_peak_subpixel

(

const hdrl_xcorrelation_result *

self

)

Get the index where the cross correlation reaches its maximum, with sub-pixel precision.

Parameters

self	hdrl_xcorrelation_result the getter will extract the data from

Returns

the index where the cross correlation reaches its maximum, with sub-pixel precision

hdrl_xcorrelation_result_get_sigma()

double hdrl_xcorrelation_result_get_sigma

(

const hdrl_xcorrelation_result *

self

)

Get the estimated standard deviation of the correlation.

Parameters

self	hdrl_xcorrelation_result the getter will extract the data from

Returns

the estimated standard deviation of the correlation

hdrl_xcorrelation_result_wrap()

hdrl_xcorrelation_result * hdrl_xcorrelation_result_wrap	(	cpl_array *	x_corr,
		const cpl_size	max_idx,
		const cpl_size	half_window
	)

Constructor for hdrl_xcorrelation_result.

Parameters

x_corr	Cross correlation. x_corr becomes owned by the returned value, do not free x_corr after the wrapping.
max_idx	Index where the cross correlation reaches its maximum
half_window	Half window used for the cross-correlation calculation

Returns

the constructed object. NULL in case of error. Errors are triggered if data are not self consistent, e.g. if max_idx is greated than the length of x_corr.