circle cells which are not so circular
3 次查看(过去 30 天)
显示 更早的评论
Hi everyone!
I am actually working on my bachelor project. I can currently detect nuclei of leucocytes, but I would like to compare number of nuclei with number of cells. How can I detect the number of leukocytes?
I tried to find circles in images with function: various radii in grayscale image via Hough Transform and imfincircles but it did not help me, I could detect number of cells just on one image and on the others it was not accurate, even when I added automatic threshold as one of the parameters for detecting circles.
I attached sample of images which I am working with and result of finding nuclei with K-means function :
![](https://www.mathworks.com/matlabcentral/answers/uploaded_files/216743/image.jpeg)
![](https://www.mathworks.com/matlabcentral/answers/uploaded_files/216744/image.jpeg)
![](https://www.mathworks.com/matlabcentral/answers/uploaded_files/216745/image.jpeg)
This last image is result of imfindcircles , but it was accurate filter just for this image, for others it detected one or more circles without detecting real cells. ![](https://www.mathworks.com/matlabcentral/answers/uploaded_files/216746/image.jpeg)
![](https://www.mathworks.com/matlabcentral/answers/uploaded_files/216746/image.jpeg)
I used:
vybrany= imread('C:\Users\Lea\Documents\Bakalarka\komprimovane\100\Screenshot_45.jpg');
imshow(vybrany), title('targetImage');
x=rgb2gray(vybrany)
imshow(x)
threshold =graythresh(x)
[centersBright,radiiBright,metricBright] = imfindcircles(x,[35 105], ...
'ObjectPolarity','dark','Sensitivity',0.96,'EdgeThreshold',(threshold./1.9));
viscircles(centersBright, radiiBright,'Color','b');
0 个评论
回答(0 个)
另请参阅
Community Treasure Hunt
Find the treasures in MATLAB Central and discover how the community can help you!
Start Hunting!