Image processing of Microscopic image

[Sena_Scott]

 Hi all,

I am working on cells manipulation therefore based on the microscopic image i have obtained from my camera. I will like to acquire the center of the particle in order to allow a laser to perform optical trapping.

At this point of time, due to the nature of the cell as shown in the attachment, the cell is surrounded by a white layer followed by pockets of white dots within the cell thus it was hard for me to obtain the exact center position of the cell.

Is there any solutions I could filter off the unncessary white dots or is there any other alternatives i could look into to get the exact center position of the cell? 

Thank you in advance. 

Best Regards,

Scott

[Todd_Lesher]

Could you fit a the white circle to a perfect circle, then use the perfect circle's geometric center?

[udka]

-Since you didn't provide proper image, it is very difficult to work with.
-Here's what i tried:
-Since it is color image, extract single plane.
-Threshold for dark objects(0-128)
-Remove Border Objects and Small Particles.
-Fill the holes and use particle analysis. There you go.

[Sena_Scott]

Todd, What do you mean by fitting a white circle to a perfect circle? 

udka, I doubt my images will be coloured. It should be in grayscale images, thus can i remove the extract single plane function? Next, when you mention threshold for dark objects (0-128), Does it means only dark objects will remain in the edited image? Lastly, by removing border objects, does it means I have to set the border via the "Particle Analysis.VI"?

Best Regards,

Scott  

[udka]

Hi Sena_Scott,
-You gave us screenshot of you image-which is color and not gray-scale (It will be nice if you share original image instead of screenshot). Since, you have gray-scale image no need of that part.
-Yes, you said you needed only cells regions which are dark in the image.
-Because of threshold for dark objects, you end-up with lot of other small particles which are connected to boundary of image and i intend to remove them.
-You need not set any border for particle analysis.

[Sena_Scott]

Hi Udka,

Sorry, my bad. I misunderstood the definition of gray-scale image. I will like to share the image however at this point of time, my program only provides a "recording function" therefore I am only showing a screenshot from the video itself. Sorry for the inconvenice caused. I will look into my program and try to implement the snapshot directly via the program. Lastly, in this sense, is it possible for me to fill up the tiny "white spots" with black to determine the center of the cell instead? 

Best Regards,

Scott

[Silver_Shaper]

---

Sena_Scott wrote:

Is there any solutions I could filter off the unncessary white dots or is there any other alternatives i could look into to get the exact center position of the cell? 

---

Dear Scott,

Yes of course, there is a solution to get the center position of the cell..

This can be done by shape detection algorithm. Select the circular shape and limit the circle radius as what you want. Look at the attached image you may need that.

[udka · Accepted solution]

Did you try Fill Holes?? I mentioned in my last step before particle analysis??

[udka]

I've little doubt with shape matching sasi, it'll be very much limited and detects the cell if it's circle only isn't it??

[Silver_Shaper]

---

@udka wrote:
I've little doubt with shape matching sasi, it'll be very much limited and detects the cell if it's circle only isn't it??

---

Dear udka,

As seen from the original image the cell is circle in shape. Author didn't mention the exact shapes.. As for as my knowledge the yeast cells are mostly circular or elliptical in nature.

In addition, there is an possiblity for cells touching together. Its hard to seperate the touching cells in binary image and processing the same.

Udka, am i answered your question??