How to measure distance in mm instead of pixel ?

I have picture with the image and scale and i would like to measure every width distance in mm rather than pixel in the attached code, (code and picture copied from @Image Analyst)

[darova]

DId you measure in pixels?

[Mehbuba Shaikh]

no

[darova]

do it

[Mehbuba Shaikh]

@Image Analyst can you edit your code now?

[Mehbuba Shaikh]

@Image Analyst can you edit your code now?

[Rik]

Why rely on others to do your work? What is stopping you from measuring the length of your legend in pixels? That way you can calculate the conversion factor between pixels and mm.

Why do you refuse to measure it yourself? If you don't know how to do it, why don't you ask for advice? You're ignoring darova's suggestion, so why shouldn't we ignore you?

[Mehbuba Shaikh]

@Rik I did measured the 200 mm by the software but i would like to develop something from my side. if you don't want to help then do not discourage the other experts. First of all understand my question, what i am asking ? or This problem may be a very small for you. Before commenting like understand the user level. Not all the user are expert like you.

@darova, sorry i missed your comment, but i want to use that 200 mm to create mm to micron factor to evaluate other dimensions. I hope you understand my question. Thanks a lot guys for asking.

[Rik]

What are the constraints of what you want to develop. Your question was how to measure the distance in mm instead of pixels. If you already have the length of that 200mm in pixels, that means you 1) already know how to measure a given distance in pixels, and 2) you already have the conversion factor (by dividing the number of measured pixels by 200).

So your question as asked is already answered. It is unclear what question you are now asking.

If somebody asks 'did you measure the distance in pixels' and the response is 'no', what do you expect the follow-up to be? Either you know how to do it (so now you do it), or you don't know how to do it (in which case you first Google it, search the documentation, or post a comment asking how to do so). Instead you just tagged Image Analyst.

Have a read here and here. It will greatly improve your chances of getting an answer.

[Mehbuba Shaikh]

Rik, apolosife for misunderstanding. I have measured this 200 mm distance through software previously. Unfortunately this software is not working in our computer now. We don't know how much pixel contain in this lines but we would like to use this scale to calculate other distances. Is it clear now ?

@Rik@Image Analyst

[David Goodmanson]

Hi Mehbuba.

How accurate would you like the result to be? You have obtained in some manner the ‘200mm’ line, which I assume is done by scaling an image with 3000x4000 pixels.

On my monitor the width of the image (to the nearest 1/2 mm) is 191.5 mm and the ‘200 mm’ line has length 27 mm. With the image being 3000 pixels wide, doing the appropriate scaling gjves (3000/191.5)*(27/200) = 2.11 pixels/mm on the original image, wherever it is.

This process has nothing to do with the actual number of pixels on my monitor, which is 1920 pixels across, or on most other monitors.

Answer 1

[Image Analyst]

Try this:

rgbImage = imread('original image3.jpg');
redChannel = rgbImage(:, :, 1);
nexttile
imshow(redChannel);
impixelinfo;
% Micron bar should always be placed by the image capture
% software at the same location for all images.  Extract it.
micronBar = mean(redChannel(3669:3672, :), 1);
nexttile
plot(micronBar, 'b-', 'LineWidth', 2);
grid on;
% Get the full width, half max of the bar
threshold = mean([min(micronBar), max(micronBar)])
yline(threshold, 'Color', 'r', 'LineWidth', 2);
leftEdge = find(micronBar <= threshold, 1, 'first');
rightEdge = find(micronBar <= threshold, 1, 'last');
% Put up vertical lines.
xline(leftEdge, 'Color', 'r', 'LineWidth', 2);
xline(rightEdge, 'Color', 'r', 'LineWidth', 2);
% Compute the width in pixels.
widthInPixels = rightEdge - leftEdge;
% Compute the spatial calibration factor in mm per pixel.
mmPerPixel = 200 / widthInPixels

threshold =

50.875

widthInPixels =

317

mmPerPixel =

0.630914826498423