I have been trying to open some RGB images, view the data as a 2D array of HSL pixels, manipulate pixels in HSL space, convert back to RGB and write manipulated image to file. However I don't quite understand how the conversions in the (awesome) julia packages Color and Images work.
For example, I expect the code below (partially written following the example from this SO question) to write something very much like this image file (as test_1.png and test_2.png):
However, the code below actually produces this much darker image instead:
How should I re-arrange the arrays or images to get the output I expect?
using Color, Images
# Download file, read it in, convert colourspace to HSL and recast as array
fname=download("https://farm9.staticflickr.com/8725/17074451907_2381037c7d_m_d.jpg")
rgb=imread(fname)
hsl=convert(Image{HSL},float32(rgb))
hslArr=reinterpret(data(hsl))
# I would like to manipulate HSL data here...
# Two ways to convert manipulated array back to HSL image
hsl_1=Image(hslArr; colorspace="HSL", colordim=1, spatialorder=["x","y"])
hsl_2=reinterpret(HSL{Float32},hslArr)
# Two ways to convert HSL image to RGB image
rgb_1=convert(Image{RGB},hsl_1)
rgb_2=convert(Array{RGB{Float32}},hsl_2)
# Finally, write images to file
imwrite(rgb_1,"test_1.png")
imwrite(rgb_2,"test_2.png")
UPDATE
Thanks to #rickhg12hs finding bug in the Color.jl module, I get expected output from code above after the following steps:
Fork the source repository for Color.jl on github
Correct the conversion function (as below), pushing changes to my fork of Color.jl
Remove the default Color.jl module that comes with julia
Install my forked package using Julia's git mechanism.
Restart julia
I haven't been able to figure out how to install a forked version of a module in parallel with a previous version, but executing the following (followed by restarting julia) should temporarily fix the bug:
Pkg.rm("Color")
Pkg.clone("https://github.com/CnrLwlss/Color.jl.git","Color")
Pkg.checkout("Color","master")
Will need to switch back to original Color module once pull request goes through.
Until Color.jl gets updated and more testing implemented/passed, you can make a single character change in Color/src/conversions.jl to most likely fix this particular issue. Change - to + on line 156.
150 # Everything to HSL
151 # -----------------
152
153 function convert{T}(::Type{HSL{T}}, c::AbstractRGB)
154 c_min = min(c.r, c.g, c.b)
155 c_max = max(c.r, c.g, c.b)
156 l = (c_max + c_min) / 2 # <-- Changed '-' to '+'
On my machine, your HSL converted bird looks great now!
Related
I'm using Octave to write a script that plots a function at different time periods. I was hoping to create an animation of the plots in order to see the changes through time.
Is there a way to save each plot for each time point, so that all plots can be combined to create this animation?
It's a bit of kludge, but you can do the following (works here with octave 4.0.0-rc2):
x = (-5:.1:5);
for p = 1:5
plot (x, x.^p)
print animation.pdf -append
endfor
im = imread ("animation.pdf", "Index", "all");
imwrite (im, "animation.gif", "DelayTime", .5)
Basically, print all your plots into a pdf, one per page. Then read the pdf's as images and print them back as gifs. This will not work on Matlab (its imread implementation can't handle pdf).
This creates an animated gif
data=rand(100,100,20); %100 by 100 and 20 frames
%data go from 0 to 1, so lets convert to 8 bit unsigned integers for saving
data=data*2^8;
data=uint8(data);
%Write the first frame to a file named animGif.gif
imwrite(data(:,:,1),'/tmp/animGif.gif','gif','writemode','overwrite',...
'LoopCount',inf,'DelayTime',0);
%Loop through and write the rest of the frames
for ii=2:size(data,3)
imwrite(data(:,:,ii),'/tmp/animGif.gif','gif','writemode','append','DelayTime',0)
end
Had to come chime in here because this was the top Google result for me when I was looking for help with this. I had issues with both answers, and some other issues, too. Notably:
For Rick T's answer, the code snippet doesn't write a plot figure, it just writes matrix data. Getting the plot window was a pain.
For carandraug's answer, writing to a PDF took a very long time and made a gigantic PDF.
On my own machine, I'm pretty sure I used apt-install to get Octave, but the getframe function I found referenced in other answers wasn't found. Turns out I had installed version 4.4, which was from 2018 (>3 years old).
I removed the old version of Octave sudo apt remove octave, then installed the new version with snap. If you try octave from a terminal without it installed it should prompt you to the snap install - be sure to include the # 6.4.0 or whatever is included in the command.
Once I had the current version installed, I got access to the getframe command, which is what lets you convert directly from a figure handle to image data - this bypasses the hackey (but previously necessary step) in #carandraug's answer where you had to write to PDF or some other image as a placeholder.
I used #RickT's answer to make my own MakeGif function, which I will share with you all here. Note that MakeGif stores the filename in a persistent variable, meaning it is retained across calls. If you change the filename it will make (or overwrite!!) the new file. If you need to overwrite the current file (i.e., running the same script multiple times and want new results) then you can use clear MakeGif between calls and that will reset the persistentFilename.
Here is the code for the MakeGif function; code to test it with is provided after this:
function MakeGif(figHandle, filename)
persistent persistentFilename = [];
if isempty(filename)
error('Can''t have an empty filename!');
endif
if ~ishandle(figHandle)
error('Call MakeGif(figHandle, filename); no valid figHandle was passed!');
endif
writeMode = 'Append';
if isempty(persistentFilename)|(filename!=persistentFilename)
persistentFilename = filename;
writeMode = 'Overwrite';
endif
imstruct = getframe(figHandle);
imwrite(imstruct.cdata, filename, 'gif', 'WriteMode',writeMode,'DelayTime',0);
endfunction
And here is the code to test the function. There's a commented-out call to clear MakeGif between the blue and green colors. If you leave it commented out it will append the green sine wave to the blue sine wave, resulting in alternating colors after every cycle - again the filename is persistent in the function. If you uncomment that call then the MakeGif function will treat the green's call as "new" and trigger the overwrite of the blue sine wave and all you'll see is green.
clear all;
time = 0:0.1:2*pi;
nSamples = numel(time);
figHandle = figure(1);
for i=1:nSamples
plot(time,sin(time + time(i)),'Color','blue');
drawnow;
MakeGif(figHandle, 'test.gif');
endfor
% Uncomment the 'clear' command below to clear the MakeGif persistent
% memory, which will trigger the green sine wave to overwrite the blue.
% Default behavior is to APPEND a green sine wave because the filename
% is the same.
%clear MakeGif;
for i=1:nSamples
plot(time,sin(time + time(i)),'Color','green');
drawnow;
MakeGif(figHandle, 'test.gif');
endfor
I spent several hours on this after being super dissatisfied with laggy screen captures so I really hope this helps someone in the future! Good luck and best wishes from the Age of Covid lol.
#Chuck thanks for that code; I've been using it to save 1500-frame GIFs of simulation output, and I find that after maybe ~500 frames the time to save the next frame to the output during the call to MakeGif starts to become ... unnerving. I guess imwrite reads and writes the entirety of the output file at each call that includes the 'WriteMode','Append' pair. At frame 1500 my output is 480Mb so that becomes untenable.
An apparent rescue for this is hinted at in the doc for Octave 7.1.0's imwrite, with the suggestion that you can pass it a 4-dimensional array and write the entire image sequence with one call. I haven't been able to make this work, though: Calling imwrite that way seems to simply write the very first image in the sequence into every frame in the output file.
I was learning how to do machine learning on mldata.org and I was watching a video on Youtube on how to use the data (https://www.youtube.com/watch?v=zY0UhXPy8fM) (2:50). Using the same data, I tried to follow exactly what he did and create a scatterplot of the dataset. However when he used the scatterplot command, it worked perfectly on his side, but I cannot do it on myside.
Can anyone explain what's wrong and what I should do?
octave:2> load banana_data.octave
octave:3> pkg load communications
octave:4> whos
Variables in the current scope:
Attr Name Size Bytes Class
==== ==== ==== ===== =====
data 2x5300 84800 double
label 1x5300 42400 double
Total is 15900 elements using 127200 bytes
octave:5> scatterplot(data, label)
error: scatterplot: real X must be a vector or a 2-column matrix
error: called from:
error: /home/anthony/octave/communications-1.2.0/scatterplot.m at line 69, column 7
The error message says it all. Your data is a 2-row matrix, and not a 2-column matrix as it should be. Just transpose it with .'.
scatterplot(data.')
I dropped the label argument since it is not compatible with the communications toolbox, either in matlab or in octave.
Update:
According to news('communications'),
The plotting functions eyediagram' andscatterplot' have improved Matlab compatibility
This may be why the behaviour is different. Be ready to find other glitches, as the octave 3.2.4 used in this course is about 5 years old.
In order to use the label, you should rather use the standard octave scatter function.
Colors could be changed by choosing another colormap.
colormap(cool(256))
scatter(data(1,:), data(2,:), 6, label, "filled")
I am trying to save an image using opencv cvSaveImage function. The problem is that I am performing a DCT on the image and then changing the coefficients that are obtained after performing the DCT, after that I am performing an inverse DCT to get back the pixel values. But this time I get the pixel values in Decimals(e.g. 254.34576). So when I save this using cvSaveImage function it discards all the values after decimals(e.g. saving 254.34576 as 254) and saves the image. Due to this my result gets affected. Please Help
"The function cvSaveImage saves the image to the specified file. The image format is chosen depending on the filename extension, see cvLoadImage. Only 8-bit single-channel or 3-channel (with 'BGR' channel order) images can be saved using this function. If the format, depth or channel order is different, use cvCvtScale and cvCvtColor to convert it before saving, or use universal cvSave to save the image to XML or YAML format."
I'd suggest investigating the cvSave function.
HOWEVER, a much easier way is to just write your own save/load functions, this would be very easy:
f = fopen("image.dat","wb");
fprintf(f,"%d%d",width,height);
for (y=0 to height)
for (x=0 to width)
fprintf(f,"%f",pixelAt(x,y));
And a corresponding mirror function for reading.
P.S. Early morning and I can't remember for the life of me if fprintf works with binary files. But you get the idea. You could use fwrite() instead.
I have a homework in which i have to convert some images to grayscale and compress them using huffman encoding. I converted them to grayscale and then i tried to compress them but i get an error. I used the code i found here.
Here is the code i'm using:
A=imread('Gray\36.png');
[symbols,p]=hist(A,unique(A))
p=p/sum(p)
[dict,avglen]=huffmandict(symbols,p)
comp=huffmanenco(A,dict)
This is the error i get. It occurs at the second line.
Error using eps
Class must be 'single' or 'double'.
Error in hist (line 90)
bins = xx + eps(xx);
What am i doing wrong?
Thanks.
P.S. how can i find the compression ratio for each image?
The problem is that when you specify the bin locations (the second input argument of 'hist'), they need to be single or double. The vector A itself does not, though. That's nice because sometimes you don't want to convert your whole dataset from an integer type to floating precision. This will fix your code:
[symbols,p]=hist(A,double(unique(A)))
Click here to see this issue is discussed more in detail.
first, try :
whos A
Seems like its type must be single or double. If not, just do A = double(A) after the imread line. Should work that way, however I'm surprised hist is not doing the conversion...
[EDIT] I have just tested it, and I am right, hist won't work in uint8, but it's okay as soon as I convert my image to double.
I was hoping to be able to generate montages using PythonMagick. The documentation seems very sparse, but I've been trying to hunt it down using the code completion part of Eclipse at least, as well as a few other questions' suggestions here on Stack Overflow. It seems that the MagickWand API has the function I am looking for, according to this:
http://www.imagemagick.org/api/MagickWand/montage_8c.html
However, I cannot seem to find it in PythonMagick. Is this simply unavailable? If so I might just ditch the rest of my PythonMagick code and rely on subprocess.call on a portable ImageMagick distribution or something like that (this program will have to be portable, and run on Windows with an easy port to Mac OS... so far I have a few other PythonMagick commands working so I'd like to keep this route going if possible).
Thanks!
Using the python imagemagick/graphicsmagick bindings helps a lot, but unfortunately not all of the functionality is there yet. I actually had the same problem with #FizxMike. I needed to use montage and then do some further operations, but saving the file on hard disk and then reloading it in a proper pgmagick object in order to do the rest of the operations and saving it again was slow.
Eventually I used the subprocess solution, but instead of saving in a file, I redirect the output in stdout. Then, I use the stdout to load the image from a pgmagick.Blob in a pgmagick.Image object and do the rest of the processing in python code.
The procedure looks like this in code:
import os
import pgmagick
import subprocess
my_files = []
# Dir with the images that you want to operate on
dir_with_images = "."
for file in os.listdir(dir_with_images):
if file.endswith(".png"):
my_files.append(os.path.join(dir_with_images, file))
montage_cmd = ['gm', 'montage']
montage_cmd.extend(my_files)
# The trick is in the next line of code. Instead of saving in a file, e.g. myimage.png
# the montaged file will just be "printed" in the stdout with 'png:-'
montage_cmd.extend(['-tile', '2x2', '-background', 'none', '-geometry', '+0+0', 'png:-'])
# Use the command line 'gm montage' since there are not python bindings for it :(
p = subprocess.Popen(montage_cmd, stdout=subprocess.PIPE, stderr=subprocess.PIPE)
# Get the stdout in a variable
stdout, stderr = p.communicate()
# Load the stdout in a python pgmagick Image object using the pgmagick.Blob
# and do the rest of the editing on python code
img = pgmagick.Image(pgmagick.Blob(stdout))
# Display the image
img.display()
geometry = pgmagick.Geometry(300, 200)
geometry.aspect(True)
# Resize the montaged image to 300x200, but keep the aspect ratio
img.scale(geometry)
# Display it again
img.display()
# And finally save it <- Only once disk access at this point.
img.write('myimage.png')
I have the same problem, even pgmagick lacks the montageImage() function needed (Magick++ montage example)
This is what I do (in a Django View):
#ImageMagick CLI is better documented anyway (-background none preserves transparency)
subprocess.call("montage -border 0 -geometry "+str(cols)+"x -tile 1x"+str(len(pages))+" "+target_path[0:len(target_path)-4]+"[0-9]*.png -background none "+target_path,shell=True)`
Not fun because I have to juggle around a bunch of files first... writing to hard disk is not the fastest thing to do, then delete the temp files.
I would much rather do it all in ram.
I am still in search of a better answer myself.