helper.leafarea: Helper function to set up LeafArea run.ij

In edwardpmorris/FAST_LeafArea: Use ImageJ to derive leaf surface area from scanned images

Description

A simple wrapper passing project specific setings to \pckgLeafArea and ImageJ.

Usage

helper.leafarea(set.directory, trim.pixel = 0, check.image = F,
  distance.pixel = 1, known.distance = 1, path.imagej = NULL)

Arguments

set.directory	Set directory that contains leaf images.
trim.pixel	Number of pixels removed from edges in the analysis.
check.image	Do you want imageJ to open and show you the analysis of each image?
distance.pixel	The number of pixels for the 'known.distance'
known.distance	The distance for which 'distance.pixel' is known
path.imagej	The path where the imageJ executable is found. For Linux this must be stated.

Details

If ImageJ is installed in one of the default directories, you do not have to specify the path to ImageJ.

/Applications/ImageJ/ <Mac>

C:\Program Files\ImageJ\ <Windows>

If the OS is Linux you do need to specify the path to ImageJ.

For example, path = "~/ImageJ/"

The number of pixels per a known distance should be supplied. For an A4 sheet scanned at 150 dpi; 58.5 px/cm Use the option check.image=T to view the auto-thresholding and particle analyis results in the GUI of ImageJ.

The amount of 'cropping' of the edges of the image is set using trim.pixel; a value of 250 is suitable for the FAST template. Probelmatic images can be manually cropped without changing the resolution, in which case trim should be adjusted accordingly.

Known bugs:

Sometimes the analysis gives errors if repeated in same session. Solution is to save results and start a new R session before re-running.

Some images cause errors. Check the resolution and dimensions of the images.

Value

A dataframe with filename and calculated 'projected surface area of element' (m2)

Examples

# If using Linux define the path to the ImageJ executable,
# if not, comment out; Leaf Area should find it.
# Recommend downloading and extracting the standalone version,
# rather than using a package manager.

# set path (Linux)
path.imagej <- "~/ImageJ"

# set path to test images directory
# A4 scans at 150 dpi, 1 cm = 58.5 px
path.image <- file.path(system.file("images"
                       , package="FASTLeafArea"),"")

# run analysis on images using ImageJ 
area <- helper.leafarea(set.directory=path.image, path.imagej=path.imagej
                       , trim.pixel=250, distance.pixel=58.5, known.distance=1
                       , check.image=F)

# write to file in working directory
write.csv(area, paste0(paste(attr(area,"standard.name")
         , attr(area,"units"), sep="_", ".csv"))
         , quote = T, na = "NA", row.names = F)

edwardpmorris/FAST_LeafArea documentation built on May 15, 2019, 11:03 p.m.