Although won't be able to answer the question properly but one thing I would like to add is that, try using the older version of Image J which is kinda more familiar and user friendly (solely an opinion) than the newer one.
the basic measure option in ImageJ is simple but probably appropiate. however, it works only grayscale images i think. so you need a conversion of your images, if RGBs. if the stain sufficiently red, and no other red-containing dye on the sample, copy red channel to a new grayscale image.
if there is not a given structure you can measure the whole area of the picture (microscopic field) but before that select background on a small proper (tissue, but negative part) portion of each pictures. i assume we speak about a tissue staining here. it looks to a collagen staining... i think the birefringence signal as more specific one would be more appropriate but detect it grayscale mode as it gives a clear mixed RGB signal according to a ref.
basic measuring method to give a number (averaged pixel intensity) using averaged signal subtracting averaged background. this simple number come from Measure plug-in. The values come from whole pictures, you need a few picture for proper statictics from each samples, over ten at least, if there is more type of samples. one important issue: do not use non-tissue area on pictures for calculation. this simple technique only compares amount of collagen between samples. a more complex method would demand measure of areas, measure of different channels, normalizing, etc if you want to measure different types of collagens.