Filling the particles-bubbles by using threshold

[saule]

Hello, I'm currently analising bubble size change by using Image J. Aftersubstracting background and increasing contrast, I have tried to adjust thethreshold. However, I have to do a lot of work to clean, and erase all noiseand rubish around the bubbles. Is there's any way of doing an inverse picture or filling up the bubble inner part, so I could avoid cleaning step.


Sample of the picture 1_t000c3.jpg

Thank you very much,
Daiva

[Alexey]

First of all you should transform your image to binary using "Binary"->"Make
Binary" . In order to increase the performance of the main step you will
need to use a filter which closes the gaps in bubbles boundaries. For this
you can use "Binary"->"Close" as a first attempt and anisotropic diffusion
for the further improvement.
And the main step: "Binary"->"Fill Holes" algorithm. It will fill inner
parts automatically.

All the best,
Alexey

saule wrote

Hello, I'm currently analising bubble size change by using Image J. Aftersubstracting background and increasing contrast, I have tried to adjust thethreshold. However, I have to do a lot of work to clean, and erase all noiseand rubish around the bubbles. Is there's any way of doing an inverse picture or filling up the bubble inner part, so I could avoid cleaning step.


Sample of the picture 1_t000c3.jpg

Thank you very much,
Daiva

[Thomas Boudier]

Hi Daiva,

This is a nice example of "filling holes" followed by a "watershed" in
the "process/binary" submenu.

Thomas

> Hello, I'm currently analising bubble size change by using Image J.
> Aftersubstracting background and increasing contrast, I have tried to adjust
> thethreshold. However, I have to do a lot of work to clean, and erase all
> noiseand rubish around the bubbles. Is there's any way of doing an inverse
> picture or filling up the bubble inner part, so I could avoid cleaning step.
>
>
> Sample of the picture  http://n2.nabble.com/file/n1482486/1_t000c3.jpg
> 1_t000c3.jpg
>
> Thank you very much,
> Daiva
>  

--
/*****************************************************/
 Thomas Boudier, MCU Université Pierre et Marie Curie
 UMR 7101 / IFR 83. Bat A 328, Campus Jussieu
 Tél : [phone]  Fax : [phone]
/*****************************************************/


 

[Michael Schmid]

Hi Daiva,

as your bubbles are almost perfectly circular, you could also have a  
look at the Hough transform, e.g. the Hough Circles plugin at the  
ImageJ web site.

Michael
________________________________________________________________


On 11 Nov 2008, at 15:36, Thomas Boudier wrote:

> Hi Daiva,
>
> This is a nice example of "filling holes" followed by a "watershed"  
> in the "process/binary" submenu.
>
> Thomas
>
>> Hello, I'm currently analising bubble size change by using Image J.
>> Aftersubstracting background and increasing contrast, I have tried  
>> to adjust
>> thethreshold. However, I have to do a lot of work to clean, and  
>> erase all
>> noiseand rubish around the bubbles. Is there's any way of doing an  
>> inverse
>> picture or filling up the bubble inner part, so I could avoid  
>> cleaning step.
>>
>>
>> Sample of the picture  http://n2.nabble.com/file/ 
>> n1482486/1_t000c3.jpg
>> 1_t000c3.jpg
>> Thank you very much,
>> Daiva
>>
>
>
> --
> /*****************************************************/
> Thomas Boudier, MCU Université Pierre et Marie Curie
> UMR 7101 / IFR 83. Bat A 328, Campus Jussieu
> Tél : [phone]  Fax : [phone]
> /*****************************************************/
>
>

[saule]

Hello everyone,

Thank you very much for all your suggestions. Probably I should fhave first given an original picture:









Master-2008-EGGW180_25OCT_t000c3.jpg
As I said before, only brakground substraction, enhace contrast and adjusted threshold was done. However, there many bubbles, that are not closed. The reason for this particular case is uneven illumination. There is any way of not "loosing" the bubbles while calculating the bubble inner radius.

Thank you again,
Daiva