HandwriterRF is designed to assist forensic document examiners by performing a statistical analysis on two handwriting samples. One or both of the samples could be from unknown writers. Two hypotheses are considered:
Hp : The two documents were written by the same writer. Hd : The two documents were written by different writers.
The statistical analysis produces a score-based likelihood ratio (SLR). An SLR greater than one, indicates that the evidence supports Hp over Hd, and the larger the SLR, the stronger the support. An SLR less than one, indicates that the evidence supports Hd over Hp, and the closer the SLR is to zero, the stronger the support.
HandwriterRF requires R and RStudio IDE.
Install the handwriterRF R package. Open RStudio, navigate to the console window, and type
install.packages("handwriterRF")Open RStudio, navigate to the console window, and load handwriterRF.
library(handwriterRF)The package includes 4 example handwriting samples from the CSAFE Handwriting Database. Compare 2 of these samples. In this case, both samples are from writer 30.
sample1 <- system.file(file.path("extdata", "w0030_s01_pWOZ_r01.png"), package = "handwriterRF")
sample2 <- system.file(file.path("extdata", "w0030_s01_pWOZ_r02.png"), package = "handwriterRF")
slr <- calculate_slr(sample1, sample2)If you would like to use your own handwriting samples, scan and save them as PNG images.
sample1 <- "path/to/your_sample1.png"
sample2 <- "path/to/your_sample2.png"
slr <- calculate_slr(sample1, sample2)The result is a data frame:
- sample1_path is the file path of the first sample.
- sample2_path is the file path of the second sample.
- docname1 is the file name of the first sample.
- docname2 is the file name of the second sample.
- score is the similarity score between the two samples.
- numerator is the numerator value of the score-based likelihood ratio. Intuitively, the larger the value the more the similarity score looks like the reference ‘same writer’ similarity scores.
- denominator is the denominator value of the score-based likelihood ratio. Intuitively, the larger the value the more the similarity score looks like the reference ‘different writers’ similarity scores.
- slr is a score-based likelihood ratio that quantifies the strength of evidence in favor of ‘same writer’ or ‘different writer.’
Display the slr data frame. We hide the file path columns here so that the data frame fits on this page.
slr %>% dplyr::select(-sample1_path, -sample2_path) docname1 docname2 score numerator denominator
1 w0030_s01_pWOZ_r01.png w0030_s01_pWOZ_r02.png 0.87 0.4582274 1e-10
slr
1 4582274302
View a verbal interpretation of the score-based likelihood ratio.
interpret_slr(slr)[1] "A score-based likelihood ratio of 4,582,274,302 means the likelihood of observing a similarity score of 0.87 if the documents were written by the same person is 4,582,274,302 times greater than the likelihood of observing this score if the documents were written by different writers."