Rosemary
Posts: 23
Joined: Sat Sep 15, 2007 8:00 pm

Re: The study of the decomposition of hair over time

In regards to my photos, I took pictures of known distances in micrometers with my microscope and from there used the digital image to get a pixel to micron ratio. I don't know if that makes sense, but I got 0.58um/pixel relationship. I calibrated the ImageJ software with this information and the results are apparently in centimeters, which I can easily convert to micrometers.However, the perimeter it gives me in centimeters turns out to be about 2 micrometers, which makes no sense since it says the area is about 707 micrometers. Do you have any suggestions?
Kelly

MelissaB
Moderator
Posts: 1055
Joined: Mon Oct 16, 2006 11:47 am

Re: The study of the decomposition of hair over time

Hmmm. Can you tell us exactly what you put in all of the fields in the 'set scale' window in Image J, and then give us an example of your results? I suspect this is just a calibration problem, although I'm not sure--that does sound odd.

Rosemary
Posts: 23
Joined: Sat Sep 15, 2007 8:00 pm

Re: The study of the decomposition of hair over time

for each of the category's in Set Scale:
Distance in pixels: 77.61
Known distance: 0.004228 cm
pixel aspect ratio: 1.0
unit of length: cm

The distance in pixels I got by drawing a line between two distinct points in which I already had drawn a line before and knew the actual distance (42.28 micrometers).
After I trace hair root one, for example, I get that the area is 1.56x10^-4 cm(1.56 micrometers) and the perimeter is 0.063 cm(630 Micrometers). Last year for my project, the average length of a hair root was about 32 micrometers and the width was about 10 micrometers. So the area and perimeter both seem wrong.
Kelly

MelissaB
Moderator
Posts: 1055
Joined: Mon Oct 16, 2006 11:47 am

Re: The study of the decomposition of hair over time

You're right, that does seem wrong. Have you tried using the magic wand instead of the trace function? The easiest way to do this is to first go to 'adjust' then 'threshold', then adjust it until all the pixels of the hair root are red--those will be black and everything else will be white. (You may need to use an image manipulation program to erase the hair itself so you only have the root.). Then you click on the hair root with the magic wand tool (the one that looks as if there is light coming from the end of a stick), and click 'measure'.

See if that gives you more reasonable values.

amyc