0

I'd like to check one feature (The lower part of the RAI) if it does appear at the top of the given images. Occasionally, there may be some noises at the left or right of the images affecting the detection such as some vertical lines or the to-be-detected image is a totally different image without the feature(RAI).

What would be the best way to achieve my expectation?

I prefer to use OpenCV. But should I use ORB,FREAK,BRISK,SURF algorithms for the detection? I'm not sure if this is the approximate case to use. I wonder if there are any other good choices and ask for help. Thanks in advance. This problem bothers me quite a long time.

The given image to be detected

user3217504
  • 105
  • 1
  • 3
  • 14
  • You could use any of those feature descriptors you mentioned. Whichever gives you the best results. That said, what is the question then? – a-Jays Oct 06 '14 at 16:59
  • I used the ORB descriptors for the detection. However, the result seems not good. – user3217504 Oct 07 '14 at 08:17
  • There are certain areas where each of these decriptors stand strong. Picking one can be a little tricky, even if you know the theory behind it. – a-Jays Oct 07 '14 at 08:29
  • I understand what you mean. And I think ORB is good enough. Now I just want to detect the feature, but don't really know how to deal with the descriptors to process further. – user3217504 Oct 07 '14 at 09:49
  • Take a look at this: http://stackoverflow.com/questions/7296915/opencv-object-matching-using-surf-descriptors-and-bruteforcematcher and the related stuff there on. – a-Jays Oct 08 '14 at 06:57
  • Thanks for your comments!I appreciate with that. – user3217504 Oct 10 '14 at 09:38

0 Answers0