Dear Fazal Urrehman,

I believe the easiest way is to set the scale of your original image, before you do any processing. Then, this information will be kept even for crops generated from that original image.

Regarding the scale and scale bars on ImageJ you might find this tutorial useful:

https://www.swarthmore.edu/NatSci/nkaplin1/scalebar.htm

This guides you through the process of calculating your pixel size in the imaging plane: for example, 1 pixel = 1 mm. However, if you do know this number beforehand you can skip those steps and directly go to the “Analyze” menu and “Set Scale”

Best

Dear Rafael

I have done all these steps as you mentioned above.my main question is that, after threshold of image the soil pores not equally distributed in the whole image area. that's why i have to select interested area (dense pore area) of pores on image by using image J rectangle option. would i use the the original scaling area? or the area which shows on image after rectangle selection of image J for percent porosity calculation of image? i am waiting you'rs valuable suggestion. thanks

Dear Fazal Urrehman,

Consider for example the cartoon attached to this answer. You can easily see that the 3 regions marked have all different pore density. Therefore, if you were to crop one of these regions, then the more appropriate number representing the properties of that ROI would be the local porosity of the ROI. That is to say, you would normalize by the size of the ROI and not by the size of the full image. On the other hand, global porosity might still be informative, it all depends on how inhomogeneous your sample is on the large scale. Further, you might want to look not only into porosity percentage, but on the number of pores per area squared and the average size of the pores, as these values are also quite important.

Best,