image.ContourDetector (version 0.1.2)

image_contour_detector: Unsupervised Smooth Contour Lines Detection in an image

Description

Unsupervised Smooth Contour Detection.

Following the a contrario approach, the starting point is defining the conditions where contours should not be detected: soft gradient regions contaminated by noise. To achieve this, low frequencies are removed from the input image. Then, contours are validated as the frontiers separating two adjacent regions, one with significantly larger values than the other. Significance is evaluated using the Mann-Whitney U test to determine whether the samples were drawn from the same distribution or not. This test makes no assumption on the distributions. The resulting algorithm is similar to the classic Marr-Hildreth edge detector, with the addition of the statistical validation step. Combined with heuristics based on the Canny and Devernay methods, an efficient algorithm is derived producing sub-pixel contours.

Usage

image_contour_detector(x, Q = 2, ...)

Value

an object of class cld which is a list with the following elements

• curves: The number of contour lines found

• contourpoints: The number of points defining the contour lines found

• data: A data.frame with columns 'x', 'y' and 'curve' giving the x/y locations for each contour curve

Arguments

x

a matrix of image pixel values in the 0-255 range.

Q

numeric value with the pixel quantization step

...

further arguments, not used yet

References

Rafael Grompone von Gioi, and Gregory Randall, Unsupervised Smooth Contour Detection, Image Processing On Line, 6 (2016), pp. 233-267. tools:::Rd_expr_doi("10.5201/ipol.2016.175")

Examples

# \dontshow{
if(require(pixmap))
{
# }
library(pixmap)
imagelocation <- system.file("extdata", "image.pgm", package="image.ContourDetector")
image         <- read.pnm(file = imagelocation, cellres = 1)
# \dontshow{
image <- image[1:100, 1:100] ## speedup to have CRAN R CMD check within 5 secs
# }
x             <- image@grey * 255
contourlines  <- image_contour_detector(x, Q = 2)
contourlines
plot(image)
plot(contourlines, add = TRUE, col = "red")
# \dontshow{
}
# End of main if statement running only if the required packages are installed
# }

##
## line_segment_detector expects a matrix as input
##  if you have a jpg/png/... convert it to pgm first or take the r/g/b channel
# \dontshow{
if(require(magick))
{
# }
library(magick)
x   <- image_read(system.file("extdata", "atomium.jpg", package="image.ContourDetector"))
x
mat <- image_data(x, channels = "gray")
mat <- as.integer(mat, transpose = TRUE)
mat <- drop(mat)
contourlines <- image_contour_detector(mat)
plot(contourlines)
# \dontshow{
}
# End of main if statement running only if the required packages are installed
# }

##
##  working with a RasterLayer
##
# \dontshow{
if(require(raster))
{
# }
# \donttest{
library(raster)
x   <- raster(system.file("extdata", "landscape.tif", package="image.ContourDetector"))

contourlines <- image_contour_detector(x)
image(x)
plot(contourlines, add = TRUE, col = "blue", lwd = 10)
# }
# \dontshow{
}
# End of main if statement running only if the required packages are installed
# }