Trainable Segmentation Weka- Freezing

imagej
segmentation
trainable-weka

#1

Hello,

Very new to this. I just downloaded Fiji and am attempting to use the trainable segmentation weka to help separate between giant vesicles and background solution. However, it keeps freezing or is unable to completely analyze the image (with no error message). It gets caught on “updating features of slice 1”. How long does it typically take to apply the first set of applied classifiers/ first training? I’m wondering if it’s something to do with my image (pasted below). I want it to recognize the difference between the vesicles (specks) and background, although this image was taken on a hemocytometer so I’m not sure if the gridlines are making it difficult to analyze.

image


#2

Good day,

this is again a problematic image because it is

  1. in JPG-format which introduces compression-artifacts
    (please post a typical raw image in the original TIF- or PNG-format)
  2. spatially badly resolved
    (Use a different magnification or camera)

Does the image contain any essential color-information?
Please tell us as well what you are going to do after the vesicle detection.

Here is a first attempt to detect the vesicles by automatic thresholding (after background-removal):
vesicles_pre

Regards

Herbie


Trainable Segmentation Weka - Measure Size of Particles?
#3

How did you get this image?


#4

Please read carefully!

[…] by automatic thresholding (after background-removal):

(The automatic threshold scheme is easy to determine and please refrain from manually setting thresholds because they aren’t objective.)

For details please study the ImageJ user guide:
https://imagej.nih.gov/ij/docs/guide/index.html


#5

Hello @Izzy_Ortiz,

Sorry to get to your questions this late. I’m interested on knowing how the Trainable Weka Segmentation gets frozen. Can you tell me the steps you followed with your image?


#6

I think it was my computer!