ExtractFourier: Extract fourier parameters from close shape

In ForensicDocument: Quantification and evaluation tools for forensic document examiners

Description

This method allows to extract fourier parameters (i.e. An and Bn parameters) from closed shapes present in images.

Usage

ExtractFourier(files, n.fourier = 8, n.samp = 128, skeletonize = TRUE,
  character_pixel = 0, output, verbose = TRUE)

Arguments

files	A vector of string, giving the images' filenames to analyse.
n.fourier	A integer, giving the number of harmonics (or fourier parameters) to extract.
n.samp	A integer, giving the number of points to subsample the closed shape.
skeletonize	A logical value, indicating if the character should be skeletonized.
character_pixel	A integer value (0 or 1), indicating which pixel value is from the character (see details).
output	A string, giving the name of the output file. See details.
verbose	A logical value, indicating if progress is to be printed on the console.

Details

The vector containing the images' filenames, the file argument, is pre-processed before the analysis. Duplicated file names and files that do not exists that are discared.

character_pixel is a integer value, either 0 or 1, it indicates which pixel value is from the character. For example, if character_pixel = 1 then pixels that have a value of 1 will correspond to the character, and pixels that have a value of 0 will correspond to the background.

Value

If output is missing, it returns a list of matrices (one for each files), containing the extracted An and Bn values. Otherwise, the results are printed in a csv file (one record or row for each image).

Author(s)

Alexandre Thiery

Examples

## Not run: 
## Example using a sample character O from Marquis et al.
## In this image, black pixels are from the character (value of 0),
## and white pixels are from the bakcground (value of 1). The
## character is not skeletonized, thus we use skeletonize = TRUE.

## We will consider a re-sampling size of the character's skeleton of
## 128 (n.samp = 128). All of the Fourier parameters will be extracted,
## thus n.fourier = n.samp / 2 = 64.
file = system.file("extdata", "fig-O.png", package = "ForensicDocument")
image = ExtractFourier(files = file, n.fourier = 64, n.samp = 128,
         skeletonize = TRUE, character_pixel = 0)

## End(Not run)

ForensicDocument documentation built on May 2, 2019, 5 p.m.