What is the difference in computational time between watershed segmentation and color threshold (in one color plane)? My images usually have one to three objects that needs to be separated from a slightly contaminated background and I need to do real time processing of the images. They are small (210 by 360 pixels) but I have to process more than 20 per second. I will appreciate very much info about different segmentation methods, in terms of computational performance and some references.
Alisa,
Check the following segmentation methods which are invariant to image size (acting on the histogram level instead of the image itself):
Method 1 & 2 (see links below).
I am not sure if they will work on your small images but it does not harm to try them.
http://se.mathworks.com/matlabcentral/fileexchange/28330-a-fast-image-segmentation-using-delaunay-triangulation
http://www.mathworks.com/matlabcentral/fileexchange/37556-a-fast-image-segmentation-using-delaunay-triangulation--revisited-
http://www.mathworks.com/matlabcentral/fileexchange/56371-zhenzhou-threshold-selection
I hope that this will help you.
@Reichl, I guess you are confusing the Big O notation with the invariance property in image processing scenario. I am not talking about time complexity here.
- Badges
- Science topic
- Similar topics
- Engineering
- Electrical Engineering
- Signal Processing
- Image Processing