roi_select: Select regions of interest
In colocr: Conduct Co-Localization Analysis of Fluorescence Microscopy Images

Description Usage Arguments Details Value Examples

Select regions of interest in an image using different morphological operations

roi_select(
  img,
  threshold,
  shrink = 5,
  grow = 5,
  fill = 5,
  clean = 5,
  tolerance = 0.1,
  n = 1
)

`img`	An object of class `cimg` or a `list` of multiple `cimg` items
`threshold`	A `numeric` to be passed to `threshold` or a `vector` of values for each image in `img`
`shrink`	A `numeric` to be passed to `shrink` or a `vector` of values for each image in `img`
`grow`	A `numeric` to be passed to `grow` or a `vector` of values for each image in `img`
`fill`	A `numeric` to be passed to `fill` or a `vector` of values for each image in `img`
`clean`	A `numeric` to be passed to `clean` or a `vector` of values for each image in `img`
`tolerance`	A `numeric` to be passed to `label` or a `vector` of values for each image in `img`
`n`	A `numeric` of the number of regions of interest or a `vector` of values for each image in `img`

The function applies several imager morphological manipulations to select the regions of interest. These include threshold which sets all values below certain cut to 0; shrink/grow for pixel set dilation and erosion; fill/clean for removing isolated regions and holes. When n is provided, the individual regions (connected components) are selected where tolerance is used to determine if two pixels belong to the same region.

A cimg. The original input img with an additional attribute label. label is a vector of integers. The labels for the selected regions of interests starts from 1 and 0 is ignored. When img is a list, a list is returned.

# load images
fl <- system.file('extdata', 'Image0001_.jpg', package = 'colocr')
img <- image_load(fl)

# choose ROI
newimg <- roi_select(img, threshold = 90)

# check the ROI labels
unique(attr(newimg, 'label'))