Loading... |
Reply to author |
Edit post |
Move post |
Delete this post |
Delete this post and replies |
Change post date |
Print post |
Permalink |
Raw mail |
Good Morning,
Thanks a lot. Your suggestions are really very hope giving and helpful. I will try them and let you know about the advancements. I will try to find a scanner otherwise I will try to use the multiple photocopy machine which have multiple functions. May be this works. So nice of you. Cordially. Kashif ZEESHAN (00 33 6 74 90 18 10) Doctorant, Biopesticide Group, Laboratoire Universitaire de Biodiversité et d'Ecologie Microbienne (LUBEM), 6 Rue de l'Université, 29334, Quimper Cedex, France De : GCH [mailto:[hidden email]] Envoyé : mardi 6 mai 2008 19:37 À : Kashif Zeeshan Objet : Re: Image analysis of aggregates in a liquid medium Kashif, Try the attached macro: makeRectangle(676, 496, 1024, 1024); run("Crop"); run("RGB Split"); close(); close(); run("Duplicate...", "title=[P1010949.JPG (red)-1]"); run("Bandpass Filter...", "filter_large=50 filter_small=2 suppress=None tolerance=5"); run("Otsu Thresholding"); I mean that instead of taking an image with your camera, you can place the petri dish on a desktop (flatbed) scanner, the same scanners that you use for scanning pictures, documents etc. Epson, Kodak, canon, there are a variety of scanners that may be suitable. Place the petri dish on the scanner glass and scan the aggregates. You can use reflection or transmission mode, with or without a background: http://images.google.no/images?hl=no&q=epson%20scanner&um=1&ie=UTF-8&sa=N&ta b=wi By "play" I mean that you will need to explore, try the different settings for finding the most suitable for your purpose. It will most probably suffice to scan the images in 256 greylevels and not in color as this is most probably unnecessary and use a lot of memory. You may also want to use the shape descriptor plugin for characterizing the agglomerates: http://www.gcsca.net/IJ/Shapes.html I hope this helps, Let me know if I can help you any further. Gary. http://www.gcsca.net On May 6, 2008, at 6:21 PM, Kashif Zeeshan wrote: Hi, Thanks a lot for your email reply. It really encourages me because I am feeling this very much difficult to do this. I did not understand your following statements sorry. "Isnt possible to scan the images in a desktop scanner instead of using a camera?" I am using an ordinary digital camera and I am taking photos like I take photos of someone. And I am not understanding how desktop scanner can help in this regard plz. "Play a bit with reflection and transmission mode" I don't know that how can I play with reflection and transmission mode? In the beginning, I was using the plugin "Nucleus counter" to count the aggregates. But it worked only one time. Then I changed the procedure. Now I am using another procedure. I am sending you the protocol which I am using and also an image to better understand the situation. The liquid medium is transparent but has a yellowish brown color. In the beginning, I was taking photos with the same transparent but colored medium but now I decant the liquid medium and then I pour again 100 ml of distilled water and then pour this one with aggregates into the Petri dishes. Your help in this regard is really very precious and valuable for me. So nice of you. Cordially. <<...>> <<...>> <<...>> Kashif ZEESHAN (00 33 6 74 90 18 10) Doctorant, Biopesticide Group, Laboratoire Universitaire de Biodiversité et d'Ecologie Microbienne (LUBEM), 6 Rue de l'Université, 29334, Quimper Cedex, France -----Message d'origine----- De : ImageJ Interest Group [mailto:[hidden email]] De la part de GCH Envoyé : mardi 6 mai 2008 18:14 À : [hidden email] Objet : Re: Image analysis of aggregates in a liquid medium Kashif, Since you are having problems with the reflection I assume that the range of greylevels (colors) varies between the images. You may want to try the IJ/Process/FFT/ Bandpass filter... to keep the structures corresponding to the aggregates. Use an upper limit corresponding to approx. the size of the aggregates. Threshold after the bandpass filtering. Isnt possible to scan the images in a desktop scanner instead of using a camera?. Scanners are good and stable devices for this purpose and if you have the aggregates in a liquid transparent medium, it shouldnt be a problem. How big are the aggregates? Play a bit with reflection and transmission mode. I hope this helps, Gary. http://www.gcsca.net On May 6, 2008, at 4:06 PM, Kashif Zeeshan wrote: > HI, > > > > I am a new user of ImageJ (I am using MacBiophotonic ImageJ > upgraded to > version 1.41a) and I am trying to practice the image analysis to > quantify > the aggregates present in liquid medium. But I am facing a lot of > problems. > Sometimes, the threshold it works and sometimes it doesn't. I try > to explain > a little bit the procedure. I pour the medium with aggregates into > a large > Petri dish and then I take photo with a dark blue background by an > ordinary > digital camera (Panasonic, DMC-LS2, 5 Mega Pixels) in default mode > without > flash. I have a problem of reflection of light. After taking photos, I > transfer them to the computer and launch ImageJ. In ImageJ, I open the > selected image and then I use the toolbar button "RGB-Merge/Split" > to have > the 3 different images (transferred into red, blue and green. I > keep the red > one for further processing. On that red image, I apply the > Segmentation > plugin "Otsu Thresholding 8bit". After that I use the command > Analyze>Analyze Particles... . It worked for one assay but not > working now. I > am blocked at this stage. Can you kindly give me some suggestion to > improve > the methodology or how I can progress with this? > > > > Cordially. > > > > Kashif ZEESHAN > > (00 33 6 74 90 18 10) > > Doctorant, Biopesticide Group, > > Laboratoire Universitaire de > > Biodiversité et d'Ecologie > > Microbienne (LUBEM), > > 6 Rue de l'Université, > > 29334, Quimper Cedex, > > France > > ... [show rest of quote] <Using ImageJ for Image Analysis of
Aggregates.doc><P1010949.JPG><P1010953.JPG> |
Free forum by Nabble | Disable Popup Ads | Edit this page |