accurately measuring rectangular shapes from microscope images

I am trying to calibrate the magnification of my imaging system using a series of rectangular shapes of a specified size. I need to measure these imaged shapes in the images produced by my camera system. Does anyone know of any tools for fitting for doing this?

Here is an example image:

Answer 1

[Image Analyst]

Threshold the image

binaryImage = grayImage > someValue;

Then fill it, sum up the image, and divide by the number of rows.

[rows, columns, numberOfColorChannels] = size(grayImage);
binaryImage = imfill(binaryImage, 'holes');
numberOfWhitePixels = sum(binaryImage(:));
meanWidth = numberOfWhitePixels  / rows

[Daniel]

Thank you for your answer. Unfortunately, there is some ambiguity in choosing the threshold value. The problem is the noise in the illuminated region. Is there any particular measures I might take to mitigate this problem?

[Image Analyst]

Well who's to say what the "right" threshold is? You are. Just pick one visually and go with it. You can use my interactive/visual thresholding app https://www.mathworks.com/matlabcentral/fileexchange/?term=authorid%3A31862&sort=downloads_desc Or you can pick a level halfway between the bright edge and the dark part near it.