Hello. I have two series of images and I want to automatically detect the corresponding pairs between these series (i.e. the image pair that depicts the same area). Is there any way to do it using only the pixel intensity values?
do IOU (Intersection over union) test. Take one image get the (xmin,ymin) (xmax,ymax) positions calculate the area, and compare these areas with your another picture. If you are doing object detection you can give some certain threshold value (usually 0.5~0.7). if IOU is greater than given threshold then your area is correct
https://opencv-python-tutroals.readthedocs.io/en/latest/py_tutorials/py_feature2d/py_surf_intro/py_surf_intro.html
https://ch.mathworks.com/help/vision/ref/detectsurffeatures.html
You can also analyze and match the pixel intensities of the neighboring pixels and create a "shape matching matrix". See articles below:
Belongie S, Malik J, Puzicha J. Shape matching and object recognition using shape contexts. Pattern Analysis and Machine Intelligence. IEEE Transactions on. 2002; 24(4):509–522.
Shah, R. G., et al. "Shape matching algorithm to validate the tracing protocol of placental chorionic surface vessel networks." Placenta 36.8 (2015): 944-946.
- Badges
- Science topic
- Similar topics
- Computer Science
- Images