Info
This question is closed. Reopen it to edit or answer.
You are now following this question
- You will see updates in your followed content feed.
- You may receive emails, depending on your communication preferences.
IS a classifier can be used to select a particular area from an image? How can i do it using KNN classifier?
1 view (last 30 days)
Show older comments
i have clustered an image. Now i want to select a particular region of the image.
0 Comments
Answers (1)
Image Analyst
on 8 Apr 2017
17 Comments
Image Analyst
on 8 Apr 2017
You have to have a set of reference points and a set of test points. Do you have that? I'm guessing not. So then you can't use knn yet.
sam CP
on 8 Apr 2017
i didn't have that reference point. What is that reference point and how can i get that reference point?
Image Analyst
on 8 Apr 2017
If you don't have a set of points that you define to be certain classes, then how can you determine if other points are close to them? I think you need to understand the concept of KNN first.
sam CP
on 8 Apr 2017
Actually i just want to select the ROI by using a classifier . KNN is not a mandatory one. Instead of KNN , how can i use SVM classifier for selecting the ROI?
Image Analyst
on 8 Apr 2017
You already did it one way with kmeans(). And like I said, I didn't think that was a good method.
For what it's worth, I'm attaching my KNN demo.
sam CP
on 8 Apr 2017
What is the training data can be used for selecting the ROI? Is that features extracted from the manually segemented tumor region?
Image Analyst
on 8 Apr 2017
You have to decide that. Maybe you say it's gray level but then everything that is the same brightness will be called a tumor simply if it's the same gray level range, regardless if it's a tumor or not.
sam CP
on 8 Apr 2017
but sometimes tumor is bright and in some cases it is dark , shape is also changed in several cases
Image Analyst
on 9 Apr 2017
There are always at least two classes. In your case they would be "tumor" and "non-tumor" (or "everything else").
sam CP
on 9 Apr 2017
Okay. You meant that i have to extract the features from manually segmented tumor region(ROI) and also from the non tumor region (background). Then this will be consider as the two classes and after that these classes are used for the training purpose. by using this two training classes , the KNN classifier will select the ROI and background from the test image. Can i proceed with this?
Walter Roberson
on 10 Apr 2017
Yes.
You mentioned that the tumor might be brighter or darker than the rest. Be sure train on both kinds and both kinds of background.
sam CP
on 10 Apr 2017
how can i extract the features of tumor section (bright or dark) and background alone , since i didn't have separated sections of both.
Image Analyst
on 10 Apr 2017
I suggest you talk with the radiologist who has hired you to do this or sponsored you to get all types of images you expect to encounter. He or she may have some images.
Walter Roberson
on 10 Apr 2017
You need some separated images to train with.
If you do not have that, then you need to use something like the Human Brain Atlas to do image registration against known normal images, so that you can recognize what part of the brain has been imaged and from what angle, so that you can do background subtraction, and then try to figure out whether the parts left over are significant.
See Also
Community Treasure Hunt
Find the treasures in MATLAB Central and discover how the community can help you!
Start Hunting!An Error Occurred
Unable to complete the action because of changes made to the page. Reload the page to see its updated state.
Select a Web Site
Choose a web site to get translated content where available and see local events and offers. Based on your location, we recommend that you select: .
You can also select a web site from the following list
How to Get Best Site Performance
Select the China site (in Chinese or English) for best site performance. Other MathWorks country sites are not optimized for visits from your location.
Americas
- América Latina (Español)
- Canada (English)
- United States (English)
Europe
- Belgium (English)
- Denmark (English)
- Deutschland (Deutsch)
- España (Español)
- Finland (English)
- France (Français)
- Ireland (English)
- Italia (Italiano)
- Luxembourg (English)
- Netherlands (English)
- Norway (English)
- Österreich (Deutsch)
- Portugal (English)
- Sweden (English)
- Switzerland
- United Kingdom (English)
Asia Pacific
- Australia (English)
- India (English)
- New Zealand (English)
- 中国
- 日本Japanese (日本語)
- 한국Korean (한국어)