ImageJ Contrast Enhancement

[@qgufo]

Craig, tutorial filter remove unblur photos .

[@antoj.5119]

Hey Craig, thanks for the content. I'm trying to segment an image that has a really low contrast, what do you suggest in this case? The pixels have an awfully close gray tone, I'm finding this very difficult to solve.

[@CraigDaly]

Hi, if none of these contrast enhancement routines work then you may have to think about some kind of machine learning to do the segmentation. The 'Weka' plugin is a good place to start. I did a video on that recently. Unfortunately, working with low contrast images (without ML) is always going to be very difficult to do using an automated method. Do you have too many images to do it by hand using the draw tools?

[@antoj.5119]

@@CraigDaly I'm so sorry, youtube never let me know you answered me. It's a couple of hundred of pics. I used your tutorial on weka and it really helped in another set of pics, but for these with lower contrast it won't work.

[@CraigDaly]

@@antoj.5119 Unfortunately, thresholding/segmentation is all about good contrast. I just spent days manually outlining some structures in a 3D EM block. Hours of laborious work but the results are stunning There are 700 slices in the block. Sometimes you might just need to do it manually if you think the end result will be worth it. C.

[@antoj.5119]

@@CraigDaly Thank you very much. I'll do it manually!

[@GameIsMyOxygen]

I am unable to find the normalize checkbox in my enhance contrast. I am using the latest version, do you have any idea as to why this is happening?

[@CraigDaly]

Hi, just back from vacation -hence the slow reply. If you are trying to normalise a colour RGB image you will need to change it to type 8-bit colour.

[@maheshbiradar2510]

Sir will you help me for some image processing, related to trichome (hair like structure on leaf) counting ? I am trying but not getting proper results.

[@CraigDaly]

Hi. I replied to the email you sent. But here (for others) is my suggestion; I have looked at the images you have sent and can only think of one suggestion that does not involve simply measuring each trichome by hand. One thing to think about is; do you need to measure the entire image? If the trichomes are regularly arranged then you could just measure (by hand) a small section and extrapolate from there. Simply using the line tool to identify and measure each hair like structure. That is slow but by far the most accurate. If you need an automated method for multiple images then I suggest you try the following. (see attached image) 1. use Fiji as it has the plugins you will need 2. Change the image to 8 bit grey scale (Image/Type / 8 bit) 3. Enhance the contrast (Process/enhance contrast) 4. Threshold the hairs (Image/adjust/Threshold) and apply 5. remove single pixels (Process/noise/despeckle) 6. Skeletonise (Process/Binary/Skeletonise) 7. Analyse/Skeleton/Analyze Skeleton. That’s my best guess for now. You will need to read the instruction for the various plugins and play with the parameters to get a better result than I have.

[@maheshbiradar2510]

Thank you sir....for your great suggestion But, we reduced image area to count trichomes, it may increase error and we will not get clear variability of different genotypes, hence we try to maximize our area of counting trichomes

[@victorcat1377]

Very usefull, thank you very much !

[@shubhanshupandey9718]

When I chose the enhanced contrast option or any other option and click apply, the range og histogram changes. Initially if the the range in from 0 to 4096, after applying the contrast or any other option it changes to 0-65536. Seems like the bit changes from 12-bit to 16-bit and I dont have a clue why is this happening. Can anyone help?

[@NavidErde]

Thx for the explaination. What I had to learn the hard way is that when you use the enhance contrast function on a stack, he ask for processing the whole stack. But he isn't actually using the data of the whole stack. He's just processing every slice of the stack separatly. So the result is not quite helpful. My problem is that my stack consists of very uneven distributed slices, i.e. some slices are very dark (mean about 100), others very bright (mean about 10000). They originate from a lightsheet microscope, so when it comes to the outside parts of the tissue its getting brighter then for the inside parts. So I have to bring up an intensity distribution matching the whole stack. Is there a way to do this in imageJ? So maybe you could add something about processing 3D images?

[@CraigDaly]

Thanks for watching. If you are using the ‘enhance contrast’ option on a z-series, there should be a checkbox for ‘use stack histogram’ this will then NOT process each slice using its own histogram. Use the orthogonal viewers to examine the result. I will do a few more 3D videos. C.

[@NavidErde]

@@CraigDaly hm probably it does what it should do. I tried it again. But it doesn't work for my stack. Especially the slices around the brainstem are way darker then at the isocortex. I guess it could happen because the amount of black pixels is about >=50% at the brainstem and

[@CraigDaly]

Difficult to make suggestions without seeing the images. I wonder if it would be worth doing equalisation of each image using a region of interest that contains the min-max values you need. If the ROI works in the same position for each slice then build a simple macro. I’m not sure, just guessing at this point without seeing the data. C.

[@liamwilson5167]

very useful thankyou!

[@mirudulaelanchezhian3620]

Hi, Can you post a video in which the fluorescence microscopy image is represented as a heatmap kind of image ie., the colour of each point in the image is now changed to some other colour based on their intensity?

[@CraigDaly]

Hi, thanks for visiting. Could you check the ImageJ walkthrough video. I cover LUTs (Look up tables) which is what I think you are asking for. Let me know if that doesn’t cover it in enough detail for you. However, I agree that an individual video on LUTs could be useful. Craig.

[@mirudulaelanchezhian3620]

@@CraigDaly yeah, I did check and used it eventually.