Hello,
I need help regarding the analysis of the following image.
I want to count the number of large bubbles and their size.
Dear Mikhail Trunov ,
There are image processing tool in matlab you can use them to identify the bubbles and sort them in size.
I am not an expert in image processing.
But I can propose some basic ideas to identify the bubbles and their size.
You can make a thresholding of the pixel illumination intensity or color.
You can calculate an average pixel intensity P(x,y)
Then the intensity greater than the threshold can be considered white say 1
The pixels smaller than than P average is made zer0.
Then you can identify the contours of the bubbles by an edge detection algorithm.
After applying this algorithm you will see the contours of the bubbles.
You can count the number of pixels in each closed circles.
Knowing the pixel area one can get the size of bubbles.
After that one can sort them using size comparators.
Best wishes
Hi there,
First of all, the image you attach behaves weirdly when opened in ImageJ (I had to save it as .tiff and then convert to RGB and then back to 8-bits, I guess this is by the .png)
Ok, so, in ImageJ:
1-Adjust B&C so basically your bubbles are white and the background is black (Image>Adjust>B&C)
2-Smooth your image (Image>Process>Smooth)
3-Threshold your image (Image>Adjust>Threshold) and set the Th more convenient for you in this case
4-Process>Binary
>Close
>Fill holes
>Watershed
5-Analyze>Analyze particles
Here you can filter by size and by circularity.
For your image I used:
Size 100-Inf
Circularity 0.25-1.00
6-Save the ROIset and impose it to your original image to see if it fits well with what you expected.
I did this roughly, there maybe many other ways to do it. You may want to play around with parameters (size, circularity...). You can also write a Macro so you don't have to go each time step by step,
Once you have the ROIset you can get any kind of information regarding size and shape of your particles (Analyze> Set measurements and in the ROI manager More>Multimeasure)
Hope it helps,
Cheers,
J
Hello,
Abdelhalim is perfectly right and the steps he mentioned can be easily done with built-in functions in Matlab. Some basic examples:
https://www.mathworks.com/matlabcentral/answers/156066-how-can-find-size-of-this-objects-in-a-binary-image
and a nice example with rice
https://www.mathworks.com/help/images/ref/bwboundaries.html
The fields of the objects can be retrieved as variable A from the function [k, A] = boundary(...).
So, although the steps explained by J. Ramirez-Franco with ImageJ seem perfect, alternatively, they can be also perfromed in the very flexible environment of Matlab (the only point I am not sure about is Circularity analysis)
- Badges
- Science topic
- Similar topics
- Computer Science
- Images