How to filter the image for blobs of specified size
3 views (last 30 days)
Show older comments
Abdullah bashanfer
on 16 Jul 2016
Edited: Image Analyst
on 17 Jul 2016
This is the code I used, but it doesn't eliminate the blobs. I want to remove the obstacles only to determine the highest middle point for the robot to navigate.
rgb=imread('obstacle_scene_1.jpg');
figure, imshow(rgb);
[im,map] = rgb2ind(rgb,5);
figure, imshow(im,map)
[X_no_dither,map]= rgb2ind(rgb,5,'nodither');
figure, imshow(X_no_dither,map);
[im,map] = rgb2ind(rgb,5);
figure, imshow(im,map)
level = graythresh(X_no_dither);
bw = im2bw(X_no_dither,level);
bw = bwareaopen(bw, 50);
figure, imshow(bw)
![](https://www.mathworks.com/matlabcentral/answers/uploaded_files/154913/image.jpeg)
![](https://www.mathworks.com/matlabcentral/answers/uploaded_files/154914/image.jpeg)
0 Comments
Accepted Answer
Image Analyst
on 16 Jul 2016
bwareaopen() removes blobs of a specified size and below. You're calling that, so that part is right. There are also related functions bwareafilt() and bwpropfilt().
Your problem is segmentation, not size filtering. You need to segment your image better. For that particular image I suggest you convert to HSV color space with rgb2hsv, then threshold the s channel to find vividly colored regions. See my demos in my File Exchange, or see the Color Thresholder on the Apps tab of the MATLAB tool ribbon.
2 Comments
Image Analyst
on 17 Jul 2016
Edited: Image Analyst
on 17 Jul 2016
Fantastic! You might also want to use
% Fill holes
binaryImage = imfill(binaryImage, 'holes');
If you don't need holes filled, then don't do it since it will slow it down, but just slightly.
I'm glad I could help, and thanks for accepting the answer.
More Answers (0)
See Also
Products
Community Treasure Hunt
Find the treasures in MATLAB Central and discover how the community can help you!
Start Hunting!