I am using imageJ to quantify fluorescent bacteria and phages. Ihave tried and deciding to use either OTSU or adaptive tresholding. Is there a good system to determine what is the best strategy? Also once, the settings for adaptive tresholding have been set, should they vary between different slides or should they remain constant? What would be the best strategy to go about accurate quantification
Hi Ojas Natarajan . Global thresholding algorithms work pretty well - some outperform others. If images are not uniformly illuminated, local thresholding algorithms may be better. If imaging conditions are the same for all your slides, I would use the same algorithm.
Thank you so much Johanna Marie Dela Cruz for your input. It means a lot. The only parameter that varies maybe the z plane across slides as the focus for each slide varies by a few micro meters. Other than that all parameters are the same. I have seen papers comparing different algorithms. Are there any specific parameters to keep in mind when choosing one
Ojas Natarajan I think that in ImageJ, the TH method always depends on the histogram of the image. There is no best thresholding approach since images vary a lot. It partly depends too on your perception of what is signal vs background.
By the way Ojas Natarajan , I have a relatively new continuing tutorial series on basic image processing and analysis on YouTube. Please check out my channel and follow me if imy videos will be helpful:
https://youtube.com/channel/UCFKPbB-h1GbuOG9gwD1tXEA. thanks.
- Badges
- Science topic
- Similar topics
- Physical Sciences
- Physical Phenomena
- Luminescence
- Fluorescence