API reference

RandomSampleConsensus Class

Class implementing the Random Sample Consensus (RANSAC) algorithm for a generic data type.

RANSAC is an iterative non-deterministic algorithm used for fitting estimation in the presence of a large number of outliers in the data.

Constructors

RandomSampleConsensus ( maxDataPoints , exclusionZone )

Initializes a new instance of the RandomSampleConsensus class.

Signature

RandomSampleConsensus(int maxDataPoints=10, int exclusionZone=2)

Parameters

maxDataPoints: int The total number of sample points.

exclusionZone: int The threshold to avoid selecting samples too close to each other.

Methods

EvaluateModelScore ( model , modelSet )	Delegate for evaluating the agreement score of a model with respect to the entire dataset. Signature delegate float EvaluateModelScore(TModel model, TModel[,] modelSet) Parameters `model: TModel` The model to be evaluated. `modelSet: TModel` The set of all generated models for comparison. Returns `delegate float` The score representing the model's agreement with the data.
FindOptimalModel ( modelGenerator , modelScorer )	Calculates the best fitting model based on the RANSAC algorithm. Signature TModel FindOptimalModel(GenerateModel modelGenerator, EvaluateModelScore modelScorer) Parameters `modelGenerator: GenerateModel` The function to generate models from data points. `modelScorer: EvaluateModelScore` The function to evaluate the agreement score of a model. Returns `TModel` The model that best fits the data according to the RANSAC algorithm.
FindOptimalModel ( modelGenerator , modelScorer , dataPointsCount )	Calculates the best fitting model based on the RANSAC algorithm. Signature TModel FindOptimalModel(GenerateModel modelGenerator, EvaluateModelScore modelScorer, int dataPointsCount) Parameters `modelGenerator: GenerateModel` The function to generate models from data points. `modelScorer: EvaluateModelScore` The function to evaluate the agreement score of a model. `dataPointsCount: int` The number of data points to evaluate. Returns `TModel` The model that best fits the data according to the RANSAC algorithm.
GenerateModel ( index1 , index2 )	Delegate for generating a model from a pair of samples, represented by their indices. Signature delegate TModel GenerateModel(int index1, int index2) Parameters `index1: int` The first sample index. `index2: int` The second sample index. Returns `delegate TModel` A model generated from the provided indices.