RANSAC.h File Reference

Perform Random Sample Consensus (RANSAC), a robust fitting method. More...

#include "std.h"

Include dependency graph for RANSAC.h:

This graph shows which files directly or indirectly include this file:

Go to the source code of this file.

Functions
void	RANSAC_linear_model (int n_samples, int n_iterations, float error_threshold, float *targets, int D, float(*samples)[D], uint16_t count, float *params, float *fit_error)
	Perform RANSAC to fit a linear model. More...
void	get_indices_without_replacement (int *indices_subset, int n_samples, int count)
	Get indices without replacement. More...
float	predict_value (float *sample, float *weights, int D, bool use_bias)
	Predict the value of a sample with linear weights. More...

Detailed Description

Perform Random Sample Consensus (RANSAC), a robust fitting method.

The concept is to select (minimal) subsets of the samples for learning a model, then determining the error on the entire data set, and selecting the best fit. In determining the error, the individual sample errors are capped, so that outliers can be identified and have only a modest influence on the results.

Read: Fischler, M. A., & Bolles, R. C. (1981). Random sample consensus: a paradigm for model fitting with applications to image analysis and automated cartography. Communications of the ACM, 24(6), 381-395.

This file depends on the function fit_linear_model in math/pprz_matrix_decomp_float.h/c

Definition in file RANSAC.h.

Function Documentation

get_indices_without_replacement()

void get_indices_without_replacement	(	int *	indices_subset,
		int	n_samples,
		int	count
	)

Get indices without replacement.

Parameters

[out]	indices_subset	This will be filled with the sampled indices
[in]	n_samples	The number of samples / indices.
[in]	count	The function will sample n_sample numbers from the range 1, 2, 3,..., count

Definition at line 160 of file RANSAC.c.

References n_samples.

Referenced by RANSAC_linear_model().

Here is the caller graph for this function:

predict_value()

float predict_value	(	float *	sample,
		float *	weights,
		int	D,
		bool	use_bias
	)

Predict the value of a sample with linear weights.

Parameters

[in]	sample	The sample vector of size D
[in]	weights	The weight vector of size D+1
[in]	D	The dimension of the sample.

Returns

The predicted value

Definition at line 136 of file RANSAC.c.

References D.

Referenced by RANSAC_linear_model().

Here is the caller graph for this function:

RANSAC_linear_model()

void RANSAC_linear_model	(	int	n_samples,
		int	n_iterations,
		float	error_threshold,
		float *	targets,
		int	D,
		float(*)	samples[D],
		uint16_t	count,
		float *	params,
		float *	fit_error
	)

Perform RANSAC to fit a linear model.

Parameters

[in]	n_samples	The number of samples to use for a single fit
[in]	n_iterations	The number of times a linear fit is performed
[in]	error_threshold	The threshold used to cap errors in the RANSAC process
[in]	targets	The target values
[in]	samples	The samples / feature vectors
[in]	D	The dimensionality of the samples
[in]	count	The number of samples
[out]	parameters*	Parameters of the linear fit
[out]	fit_error*	Total error of the fit
[in]	n_samples	The number of samples to use for a single fit
[in]	n_iterations	The number of times a linear fit is performed
[in]	error_threshold	The threshold used to cap errors in the RANSAC process
[in]	targets	The target values
[in]	samples	The samples / feature vectors
[in]	D	The dimensionality of the samples
[in]	count	The number of samples
[out]	parameters*	Parameters of the linear fit, of size D + 1 (accounting for a constant 1 being added to the samples to represent a potential bias)
[out]	fit_error*	Total error of the fit

Definition at line 57 of file RANSAC.c.

References D, fit_linear_model(), fit_linear_model_prior(), get_indices_without_replacement(), n_samples, and predict_value().

Here is the call graph for this function: