Some papers use image segmentation techniques and others use edge detection. I need to know which one is faster.


I need to do semantic segmentation for outdoor sense and the data is optical image and LiDAR. Other researchers use segmentation before doing feature detection but I expect that if I use line/edge detection, the technique will detect image features faster.

Before I start doing experiments I want to know which one is better because this will effect on next step, i.e. feature detect.

This application should work in real-time and I plan to use GPU for faster processing, such as CUDA.

  • $\begingroup$ Welcome to SP.SE! Your question is very, very general and not clear. You need to provide much more information: what sort of images are you using? What is the application that you need edge detection / segmentation for? What have you tried? How fast do you need it? Closing until we get some more details. $\endgroup$
    – Peter K.
    Oct 13, 2016 at 16:54
  • $\begingroup$ @peterK. I add all info you ask for $\endgroup$ Oct 17, 2016 at 0:30

1 Answer 1


Image segmentation is way harder and more complex than a local operation like edge detection. You may do things in parallel for edge detection as one does not interfere with the other. Global segmentation can't allow this.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.