R/imageArea.R

In rossholmberg/holmberg: A personal package by Ross Holmberg

#' Calculate the area of an image with colour EXCEEDING threshold grey value
#'
#'
#' @param file A jpeg file to analyse
#' @param thresholdGrey Numeric value 0-1, threshold greyscale value
#' @param which Either "high" or "low", specifying what type of thresholding to perform
#' @keywords image analysis jpeg
#' @export
#' @import jpeg
#' @return A numeric value, pixel count of matching colour


areaThreshold <- function( file, thresholdGrey = 0.5, which = "high" ) {
    
    # read in the file
    input <- jpeg::readJPEG( file )
    
    input <- mean( input[,,1], input[,,2], input[,,3] )
    
    if( which == "high" ) {
        
        # mark which pixels exceed the threshold grey value
        input <- 1 * ( input[,,1] > thresholdGrey )
    } else if( which == "low" ) {
        
        # or fall below it
        input <- 1 * ( input[,,1] <= thresholdGrey )
    } else {
        stop( "\"which\" argument must be either \"high\" or \"low\"" )
    }
    
    # calculate the area
    return( sum( input ) )
}



#' Calculate the area of an image matching a specific colour
#'
#'
#' @param file A jpeg file to analyse
#' @param colour Numeric vector of colour components, c( red, green, blue )
#' @keywords image analysis jpeg
#' @export
#' @return A numeric value, pixel count of matching colour


areaColoured <- function( file, colour = c( 0, 0, 0 ) ) {
    
    # read in the file
    input <- jpeg::readJPEG( file )
    
    # mark which pixels match the colour we're searching for
    input <- 1 * ( input[,,1] == colour[1] & 
                       input[,,2] == colour[2] & 
                       input[,,3] == colour[3] )
    
    # calculate the area
    return( sum( input ) )
}