Gen Art Manual
User Manual:
Open the PDF directly: View PDF 
.
Page Count: 9
Package ‘GenArt’
June 4, 2019
Title Genetic algorithm for image reproduction
Version 0.1.0
Description R package provides a genetic algorithm for image reproduction. It is based on forc-
ing one or more errors during DNA replication (mutation) to keep the best candidate for an-
other mutation. DNA consists of coordinates, degree of trans-
parency, and RGB codes or gray scales for a given number of polygons.
Depends R (>= 3.5.1)
License MIT + file LICENSE
Encoding UTF-8
LazyData true
Maintainer Robert Herrmann <robertherrmann@mail.de>
Author Robert Herrmann
RoxygenNote 6.1.1
Import png
NeedsCompilation no
Rtopics documented:
dna_convert......................................... 2
dna_in............................................ 2
dna_mutate ......................................... 3
dna_print .......................................... 3
dna_read........................................... 4
dna_tangle.......................................... 4
dna_untangle ........................................ 5
gen_art ........................................... 5
pic_compare......................................... 6
pic_convert ......................................... 6
pic_randomtiles....................................... 7
pic_unitiles ......................................... 7
Index 9
1
2dna_in
dna_convert Convert image DNA to PNG format
Description
Function converts image DNA to array object including RGB or gray scale for each pixel.
Usage
dna_convert(dna, maxXY, tempf, pngWH, bg = "white")
Arguments
dna matrix or character, untangled or tangled image DNA of any size.
tempf temporate file generated by default or given as file path.
pngWH vector, width and height of reconstructed image. If missing, width and height of
original image are used.
bg character, color or RGB code indicating the background color of PNG.
Details
See example...
Examples
dna <- dna_untangle(dna_in(rgb = FALSE))
for(i in 1:20){
dna <- dna_mutate(dna)
}
test <- dna_convert(dna)
grid::grid.raster(test)
test[1,1,]
dna_in Generate image DNA
Description
Function to generate a random image DNA.
Usage
dna_in(n = 10, vertex = 3, maxXY = c(200, 200), rgb = TRUE)
Arguments
nnumeric, number of polygons to be drawn to your image, Default = 10.
vertex numeric, number of vertices each polygen should be constructed of, Default =
3.
maxXY vector, sets the pixels given by the original image, Default = c(200, 200).
rgb logic, if TRUE, generates an RGB code (default), if FALLS, generates a gray
scale.
dna_mutate 3
Details
Is generally the first step of starting an image reproduction. Use maxXY = dim(pic)[2:1] to get the
number of pixels in the original image.
Examples
dna_in(n = 5, vertex = 2, rgb = FALSE)
dna_mutate Mutate image DNA
Description
Function to mutate an image DNA.
Usage
dna_mutate(dna, degree = "soft", maxXY)
Arguments
dna matrix or character, untangled or tangled image DNA of any size.
degree numeric, number nucleotides to be modified per mutation, Default = 1.
Details
It is recommended to use the softest mutation rate as given by default.
Examples
dna <- dna_untangle(dna_in(rgb = FALSE))
test <- dna_mutate(dna, degree = 20)
pic_compare(dna, test)
dna_print Plot image DNA
Description
Function to plot an image DNA.
Usage
dna_print(dna, maxXY)
Arguments
dna matrix or character, untangled or tangled image DNA of any size.
4dna_tangle
Details
Are comming up soon...
Examples
#Is comming up soon...
dna_read Read image DNA
Description
Function to read file.txt including a tangled or untangled image DNA.
Usage
dna_read(log, tangle = TRUE)
Arguments
log filepath.
tangle logic, TRUE if tangled (default), FALSE if untangled.
Details
After your calculation is done, you may want to improve the results by reloading the existing image
DNA, and start the calculation again.
dna_tangle Tangle image DNA
Description
Function to tangle an untangled image DNA.
Usage
dna_tangle(dna)
Arguments
dna matrix, untangled image DNA of any size.
Details
See example...
Examples
dna <- dna_untangle(dna_in(n = 2, vertex = 3, rgb = FALSE))
dna_tangle(dna)
dna_untangle 5
dna_untangle Untangle image DNA
Description
Function to untangle an image DNA to get color code and coordinates of each polygon.
Usage
dna_untangle(dna)
Arguments
dna character, tangled image DNA of any size.
Details
See example...
Examples
dna_untangle(dna_in())
dna_untangle(dna_in(rgb = FALSE))
gen_art Genetic algorithm
Description
Algorithm that continues to mutate (improve) the image DNA until manually stoped or a given
maximum of iterations is reached
Usage
gen_art(pic, dna, iter = Inf, fname = "ReconPic.png",
fnamelog = "Log.txt", tempf, degree = "Soft")
Arguments
pic object of class array, original image in PNG format.
dna matrix or character, untangled or tangled image DNA of any size.
iter numeric, number of iterations. Default = Inf.
fname character, filename for the reconstructed image. Default = "ReconPic.png".
tempf temporate file generated by default or given as file path. See ...
degree numeric, how many nucleotides should be modified per one iteration, Default =
1.
fanmelog character, filename for file.txt including the tangled image DNA. Default =
"Log.txt".
6pic_convert
Details
It is recommended to use the softest mutation rate as given by default.
See example...
Examples
#Is coming up soon...
pic_compare Calculate image fitness
Description
Function to compare two images to calculate fitness.
Usage
pic_compare(imgONE, imgTWO)
Arguments
imgONE first array to compare with second array.
imgTWO second array to compare with first array.
Details
...
pic_convert Convert RGB code to gray scale
Description
Function to convert RGB to gray scale.
Usage
pic_convert(pic)
Arguments
pic array including RGB.
Details
...
pic_randomtiles 7
pic_randomtiles Randomly crop an image
Description
Function to tile an image into 4 different pieces.
Usage
pic_randomtiles(w, h, plot = TRUE)
Arguments
wnumeric, width of the original picture.
hnumeric, height of the original picture.
plot logic, function plots layout if TRUE (default).
Details
Try multiple times to get the required image layout. Assign data frame to an object for subsequent
use in gen_art. See example.
Examples
## Image width and height
h <- 100
w <- 120
## Three trials as an example
for(i in 1:3){
pic_randomtiles(w, h, plot = TRUE)
title("Split your image for better results...")
text(w/2, h/2, paste("Trial", i), cex = 4, col = "white")
Sys.sleep(1.5)
}
pic_unitiles Uniformly crop an image
Description
Function to uniformly crop an image into n pieces.
Usage
pic_unitiles(wh, nn, plot = TRUE)
8pic_unitiles
Arguments
wh vector, width and height of the original image.
nn vector, number of columns and rows to split the image in.
plot logic, function plots layout if TRUE (default).
Details
Function cuts remaining pixels if width/height is indivisible by columns/rows. Assign data frame
to an object for subsequent use in gen_art. See example.
Examples
## Image width/height and number of colums/rows
wh <- c(89, 70)
nn <- c(6, 4)
## Calculation and plotting
pic_unitiles(wh, nn)
