can I get some idea from the expert of MATLAB?

I want to use 4 Basler Ace cameras to capture images using MATLAB. How can I start writting MATLAB script for this.

Any help will be appreciated!

[Walter Roberson]

USB-3, GigE, or CameraLink interface?

What resolution and frame rate are you hoping for? RGB, YCrCb (typically compressed compared to rgb), grayscale?

[Chunru]

Download the webcam support package and follow some examples there.

https://www.mathworks.com/matlabcentral/fileexchange/45182-matlab-support-package-for-usb-webcams#:~:text=MATLAB%C2%AE%20Support%20Package%20for,your%20computer%20via%20USB%20port.

If you have DNN toolbox, try the example: Try Deep Learning in 10 Lines of MATLAB Code

https://www.mathworks.com/help/deeplearning/gs/try-deep-learning-in-10-lines-of-matlab-code.html

[Raushan]

@Walter Roberson I am using Basler ace USB-3. Resolution and fps can be changed right!

[Raushan]

@Chunru Thanks for the suggestions! I have the support package installed.

if I want to use this script for basler ace camera then will it work? Because it is saying to install webcam.

[Chunru]

Then you can try the code:

camera = webcam; % Connect to the camera
while true
    im = snapshot(camera);       % Take a picture
    image(im);                   % Show the picture
    drawnow
    pause(0.1)                   % you can remove this
end

[Walter Roberson]

https://www.mathworks.com/products/image-acquisition.html Image Acquisition Toolbox might be what you want.

[Raushan]

I tried Image acquisition app in matlab for 1 camera and got preview and captured some image too. Here is the code for capturing one image from one camera

vid = videoinput('gentl', 2, 'Mono8');

src = getselectedsource(vid);

vid.FramesPerTrigger = 1;

vid.ROIPosition = [1273 253 806 909];

start(vid);

stoppreview(vid);

imwrite(getdata(vid), 'E:\Raushan\7-19-22\pic_5.tiff');

I want to use 4 basler ace camera camera to captured 4 image at the same time. Can you please help me to proceed with. I will appriciate any kind of help.

Answer 1

[Image Analyst]

Machine vision cameras can often be seen as generic webcams, though not with all the functionality or resolutions you'd get with the Image Acquisition Toolbox. For example, see my camera list below.

I have three cameras: (1) integrated webcam inside the laptop computer, (2) external Logitech c930e webcam, and (3) external Lumenera machine vision camera.

The upper list is a list of all possible resolution modes for the Lumenera camera that are available with the Image Acquisition Toolbox. You can see there are a lot. The camera uses its own third party (manufacturer) adapter called lumeneraimaq-1. It does not use a winvideo adapter in that case.

However the Lumenera camera also shows up at the bottom as winvideo-3, but it has only two resolution modes to choose from. And of course a lot of other functionality that the camera has is missing, like flipping/mirroring, white balancing, adjusting gains of each color channel independently, frame rate, etc. If I want to use the Lumenera camera as a dumb webcam, I can select the winvideo-3 adapter for it and I believe it would still work. I haven't tried it because why would I want to give up all the functionality I have available to me?

So you might be able to run your Basler camera as just a dumb webcam if you want.

https://www.mathworks.com/videos/deep-learning-with-matlab-transfer-learning-in-10-lines-of-matlab-code-1487714838381.html?s_tid=srchtitle_Deep%20Learning%20in%2010%20Lines%20of%20MATLAB%20Code_1

https://www.mathworks.com/help/deeplearning/gs/try-deep-learning-in-10-lines-of-matlab-code.html?s_tid=srchtitle_Deep%2520Learning%2520in%252010%2520Lines%2520of%2520MATLAB%2520Code_2

camera = webcam; % Connect to the camera
net = alexnet;   % Load the neural network
while true
    im = snapshot(camera);       % Take a picture
    image(im);                   % Show the picture
    im = imresize(im,[227 227]); % Resize the picture for alexnet
    label = classify(net,im);    % Classify the picture
    title(char(label));          % Show the class label
    drawnow
end

[Raushan]

Thanks for the suggestion but I don't have knowledge about neural network.

what I am trying, I used MATLAB app for image acquisition to captured 1 image from 1 basler camera and it worked. But I want to use 4 Basler camera to capture 4 image at the same time. can you please give some idea how do I proceed?

Thanks in advance!

[Image Analyst]

That was just a fun example. You don't need that. You could simply do

camera = webcam; % Connect to the camera
frameCount = 0;
while true
    im = snapshot(camera);       % Take a picture
    image(im);                   % Show the picture
    frameCount = frameCount + 1; % Increment the frame counter.
    caption = sprintf('Frame #%d'); % Prepare a title for the image.
    title(caption);              % Show the label
    drawnow
end

[Raushan]

Thanks for your valuble suggetsion but i did something like this

vid = videoinput('gentl', 1, 'Mono8');
vid1 = videoinput('gentl', 2, 'Mono8');
vid2 = videoinput('gentl', 3, 'Mono8');
frameCount = 0;
while true
    src = getselectedsource(vid,vid1,vid2);
    frameCount = frameCount + 1;
    vid.ROIPosition = [1273 253 806 909];
    start(vid);
    stoppreview(vid);
    imwrite(getdata(vid), 'E:\Raushan\7-19-22\pic_5.tiff');
end

it shows following error

Error using imaqdevice/getselectedsource

Too many input arguments.

please help me out.

Thanks a lot!

[Raushan]

When I am running the following code it takes three image from three cameras but when I am trying to take simultaneously it shows error

vid = videoinput('gentl', 1, 'Mono8');
vid1 = videoinput('gentl', 2, 'Mono8');
vid2 = videoinput('gentl', 3, 'Mono8');


src = getselectedsource(vid);
src = getselectedsource(vid1);
src = getselectedsource(vid2);


vid.FramesPerTrigger = 1;
vid1.FramesPerTrigger = 1;
vid2.FramesPerTrigger = 1;



vid.ROIPosition = [1273 253 806 909];
vid1.ROIPosition = [1273 253 806 909];
vid2.ROIPosition = [1273 253 806 909];

start(vid);
stoppreview(vid);
imwrite(getdata(vid), 'E:\Raushan\7-19-22\pic_10.tiff');

start(vid1);
stoppreview(vid1);
imwrite(getdata(vid1), 'E:\Raushan\7-19-22\pic_11.tiff');

start(vid2);
stoppreview(vid2);
imwrite(getdata(vid2), 'E:\Raushan\7-19-22\pic_12.tiff');

can you please give some idea how do I put these codes for simultaneous image. Thanks a lot!

[Walter Roberson]

The following is untested, as I do not have your hardware.

outdir = 'E:\Raushan\7-19-22';
outprefix1 = "pic1_";  %string, not character vector!
outprefix2 = "pic2_";
outprefix3 = "pic3_";
img_suffix = ".tiff";

FramesToFetch = 20;

vid1 = videoinput('gentl', 1, 'Mono8');
vid2 = videoinput('gentl', 2, 'Mono8');
vid3 = videoinput('gentl', 3, 'Mono8');


src1 = getselectedsource(vid1);
src2 = getselectedsource(vid2);
src3 = getselectedsource(vid3);

%I cannot tell at the moment whether FramesPerTrigger needs to be set
%against the vid* or the src* with the hints tending to favour src*
src1.FramesPerTrigger = 1;
src2.FramesPerTrigger = 1;
src3.FramesPerTrigger = 1;

%I cannot tell at the moment whether ROIPosition needs to be set
%against the vid* or the src* with the hints tending to favour vid*
vid.ROIPosition = [1273 253 806 909];
vid1.ROIPosition = [1273 253 806 909];
vid2.ROIPosition = [1273 253 806 909];


start([vid1, vid2, vid3]);

for framecount = 1 : FrameToFetch
    
  outname1 = fullfile(outdir, outprefix1 + framecount + img_suffix);
  outname2 = fullfile(outdir, outprefix2 + framecount + img_suffix);
  outname3 = fullfile(outdir, outprefix3 + framecount + img_suffix);
  imwrite(getdata(vid1), outname1);
  imwrite(getdata(vid2), outname2);
  imwrite(getdata(vid3), outname3);
end

stop([vid1, vid2, vid3]);

[Raushan]

FramesPerTrigger needs to be set against the vid.

ROIPosition needs to be set against the vid. When I run this code I got three image pic1.1, pic2.1, and pic3.1 with following error

Error using imaqdevice/getdata (line 154)

OBJ is not running and no frames are available.

Thank you very much! you are awesome. You guys are helping selflessly, I really appreciate that!

My email- in case if you need any help from my side :)

[e-mail]

[Walter Roberson]

outdir = 'E:\Raushan\7-19-22';
outprefix1 = "pic1_";  %string, not character vector!
outprefix2 = "pic2_";
outprefix3 = "pic3_";
img_suffix = ".tiff";

FramesToFetch = 20;

vid1 = videoinput('gentl', 1, 'Mono8');
vid2 = videoinput('gentl', 2, 'Mono8');
vid3 = videoinput('gentl', 3, 'Mono8');


src1 = getselectedsource(vid1);
src2 = getselectedsource(vid2);
src3 = getselectedsource(vid3);

%I cannot tell at the moment whether FramesPerTrigger needs to be set
%against the vid* or the src* with the hints tending to favour src*
src1.FramesPerTrigger = 1;
src2.FramesPerTrigger = 1;
src3.FramesPerTrigger = 1;

%I cannot tell at the moment whether ROIPosition needs to be set
%against the vid* or the src* with the hints tending to favour vid*
vid.ROIPosition = [1273 253 806 909];
vid1.ROIPosition = [1273 253 806 909];
vid2.ROIPosition = [1273 253 806 909];


vidlist = [vid1, vid2, vid3];
start(vidlist);

for framecount = 1 : FrameToFetch
  trigger(vidlist);
  
  outname1 = fullfile(outdir, outprefix1 + framecount + img_suffix);
  outname2 = fullfile(outdir, outprefix2 + framecount + img_suffix);
  outname3 = fullfile(outdir, outprefix3 + framecount + img_suffix);
  imwrite(getdata(vid1), outname1);
  imwrite(getdata(vid2), outname2);
  imwrite(getdata(vid3), outname3);
end

stop([vid1, vid2, vid3]);

Note that over the long term it would probably be better to acquire more frames per trigger. You can use getdata(vid1, 1) to fetch the frame from the buffers of vid1 even if there are multiple frames already acquired

[Raushan]

@Walter Roberson, The edited code about which I emailed you.

outdir = 'E:\Raushan\7-19-22';

outprefix1 = "pic1_"; %string, not character vector!

outprefix2 = "pic2_";

outprefix3 = "pic3_";

img_suffix = ".tiff";

FramesToFetch = 20;

vid1 = videoinput('gentl', 1, 'Mono8');

vid2 = videoinput('gentl', 2, 'Mono8');

vid3 = videoinput('gentl', 3, 'Mono8');

src1 = getselectedsource(vid1);

src2 = getselectedsource(vid2);

src3 = getselectedsource(vid3);

%I cannot tell at the moment whether FramesPerTrigger needs to be set

%against the vid* or the src* with the hints tending to favour src*

src1.FramesPerTrigger = 1;

src2.FramesPerTrigger = 1;

src3.FramesPerTrigger = 1;

%I cannot tell at the moment whether ROIPosition needs to be set

%against the vid* or the src* with the hints tending to favour vid*

vid.ROIPosition = [1273 253 806 909];

vid1.ROIPosition = [1273 253 806 909];

vid2.ROIPosition = [1273 253 806 909];

vidlist = [vid1, vid2, vid3];

start(vidlist);

pics=cell(1,n)

for framecount = 1 : FrameToFetch

trigger(vidlist);

for i=1:n

pause(600); %After every 600sec

outname1 = fullfile(outdir, outprefix1 + framecount + img_suffix);

outname2 = fullfile(outdir, outprefix2 + framecount + img_suffix);

outname3 = fullfile(outdir, outprefix3 + framecount + img_suffix);

imwrite(getdata(vid1), outname1);

imwrite(getdata(vid2), outname2);

imwrite(getdata(vid3), outname3);

end

end

stop([vid1, vid2, vid3]);

Thank you so Much!

[Walter Roberson]

What is n? Why are you writing to the same files n times?

[Raushan]

I want to keep taking images till the end condition meets.

[Raushan]

To give end condition I want to implement these: First time when the cameras will take images we will have 3 images then I want to say to find “y” coordinates of these images. Now, when the cameras will take another 3 images after 10min then I want to find the difference in Y coordinates of new images and old images . And if there is difference in “y” coordinates then I want to say keep taking images. And if the difference is too high I want say stop taking images.

Please help me out to implement this.

Thank you very much!

[Walter Roberson]

Take 3 initial images. 
Make them the old images 

while true
  wait 10 minutes
  trigger
  record the 3 images
  compare the images to the old images 
  if they are different enough, break 
  make the images just recorded the old images
  end of loop