I have worked with acoustic image processing in a harbor and, after the preprocessing step, I need to extract track features between these images. Following below two examples of preprocessed frames:
Frame 1:
https://i.stack.imgur.com/Kuffj.png
Frame 2:
https://i.stack.imgur.com/H8LhV.png
My questions:
- What is the best feature extractor and descriptor I can use here, since the objects present are similar (pillars) and there is no much information in their neighborhood?
- How can I track these features with high success rate?
Regards,
RĂ´mulo
--- SoupGate-Win32 v1.05
* Origin: fsxNet Usenet Gateway (21:1/5)