I was unable to set threshold for this picture (TEM image). The circular black spots the nanoparticles and I have to determine the size of these particles. When I set threshold and analyze particles, Image J splits up these spots into a large number of dots and analyses. This may affect the particle size. This does not give actual particle size. My TEM image has large no. of nano particles many of which are overlapped. I find Image J useful for the images with less noise and separate particles. Can anyone help me to set threshold limit for the TEM images?
Unfortunately, if the particles overlap, ImageJ can't autocount them. There's currently a great deal of computer science academics working on developing programmes that can do this e.g. for cell counting, but it is incredibly difficult to get the computer to be able to tell where an edge is in overlapping particles. My particles are magnetic and always overlap, so I nearly always have to manually measure the length and width of each individual particle. Unless you can re-do your TEM grids so the particles are less densely packed and re-image them, you'll probably have to manually measure them. I can process about 4-500 particles infront of the tv in about 1-2 hours, as it doesn't take much brain power once you get the hang of it.
Unfortunately, if the particles overlap, ImageJ can't autocount them. There's currently a great deal of computer science academics working on developing programmes that can do this e.g. for cell counting, but it is incredibly difficult to get the computer to be able to tell where an edge is in overlapping particles. My particles are magnetic and always overlap, so I nearly always have to manually measure the length and width of each individual particle. Unless you can re-do your TEM grids so the particles are less densely packed and re-image them, you'll probably have to manually measure them. I can process about 4-500 particles infront of the tv in about 1-2 hours, as it doesn't take much brain power once you get the hang of it.
Hello,
As Johanna Galloway says, if the particles overlap, it will be very difficult to count them, also if you use an other software.
In fact, I have the same problem as you with my SEM images. Some simulations have to be made on the interaction of electrons with a nanoparticle to know the "real" threshold to apply to images.
I know that nanometrolgy team from PTB (Metrology Institute from Germany) works on the subject and I give you a link to an article from this team which might help you:
https://www.researchgate.net/publication/224041576_Characterization_of_nanoparticles_by_scanning_electron_microscopy_in_transmission_mode?ev=srch_pub
Article Characterization of nanoparticles by scanning electron micro...
Manually and directly you could make thresholds. Take a hard copy print out and with the help of scale you can interpret the image.
Well first of all make more dilute sample on the TEM grid. This will prevent a bit of overlapping. Second with ImageJ you can select the area you want to work on, i.e. the overlapped particles.
Hi Deepak Sapkota
I agree with Johanna. When the particles overlap on each others, there will no software that is able to credibly analyse the particle morphology,. My suggestion is firstly try to manually "separate" the particles (on your images) from each others, by marking the particles and tracing over their edge, one by one. Next, you convert your manual work (now is a pictures of many many individual particles) to a format (jpg, bmp, tif...) that can be analyzed by ImageJ, and then just do the analysis like usual.
Regards
One point to consider is that if you choose software that tries to fit the particle diameters in partially overlapping particles, then
a) this practically always assumes the particles are round, which at the atomic level is probably wrong (as most particles
- Badges
- Science topic
- Similar topics
- Chemistry
- Nanotechnology
- Nanostructured Materials
- Nanoparticles