segment_elan_image: Segment image(s) from video based on ELAN annotations

In borstell/signglossR: Facilitates Visual Representation of Sign Language Data

Segment image(s) from video based on ELAN annotations

Description

This function inputs an ELAN file and a video and outputs images from the ELAN annotations as segments

Usage

segment_elan_image(
  eaf,
  segmentation_tier = NULL,
  gloss_tier = NULL,
  target_glosses = NULL,
  frame_offset = 0,
  gloss_text = F,
  video = NULL,
  image_format = ".jpg",
  path = NULL,
  tag = T,
  trim = 1,
  scale = 1,
  crop_region = NULL,
  text = NULL,
  textcolor = "white",
  boxcolor = NULL,
  font = "Helvetica",
  fontsize = "70",
  rel_fontsize = NULL,
  gravity = "southwest",
  location = "+20+20",
  border = FALSE
)

Arguments

eaf	The ELAN files that supplies annotations as segments
segmentation_tier	Specify tier (TIER_ID) for segments
gloss_tier	Specify tier(s) (TIER_ID) for glosses
target_glosses	Specify which glosses to be kept (default is all glosses)
frame_offset	Select image +/-n from the gloss annotation's start frame (default = 0)
gloss_text	Whether to include text from 'gloss_tier' (default = 'FALSE')
video	The filename of the video to be modified
image_format	The format of the output images (default = '.jpg')
path	The path to the output images
tag	Whether to use annotations as tags in filenames (default = 'TRUE')
trim	Optional argument for horizontal image trim (0 to 1; default = 1)
scale	The resizing proportion of the original image (0 to 1; default is 1 = original size)
crop_region	The region to be cropped (e.g. '100x150+100+100', (width x height +upper_x +upper_y)).
text	The text string to be added to the image
textcolor	The color of the text annotation (default is "white")
boxcolor	The color of the textbox (default is 'NULL')
font	The font of the text annotation (default is "Helvetica")
fontsize	The fontsize of the text annotation (default is 70)
rel_fontsize	Whether to adjust the fontsize relative to width (x) or height (y)
gravity	The anchor point of the added text (default is "southwest")
location	The location of the text relative to the anchor point
border	Set to 'TRUE' if you want a black border frame for the output image

Value

The names of the output image files

borstell/signglossR documentation built on June 11, 2025, 5:58 a.m.