Can someone give a detailed explanation IoU and IoBB along with that the differences between them.

  • 1
    $\begingroup$ Can you give a reference that uses "intersection over bounding box"? Googling the exact term returns no result. $\endgroup$
    – user12075
    Sep 28, 2018 at 22:48
  • $\begingroup$ The two answers here have nice explanations. There isn't a big difference between the two. We are strictly talking about bounding boxes in the latter, but it uses the same principle. $\endgroup$
    – n1k31t4
    Sep 29, 2018 at 5:52
  • $\begingroup$ For reference, you can see the following paper: title={Chestx-ray8: Hospital-scale chest x-ray database and benchmarks on weakly-supervised classification and localization of common thorax diseases}, author={Wang, Xiaosong and Peng, Yifan and Lu, Le and Lu, Zhiyong and Bagheri, Mohammadhadi and Summers, Ronald M}, booktitle={Computer Vision and Pattern Recognition (CVPR), 2017 IEEE Conference on}, pages={3462--3471}, year={2017}, organization={IEEE} $\endgroup$ Oct 2, 2018 at 12:37

2 Answers 2


The Intersection over Bounding Box is the Intersection over Union (IoU) for object detection tasks, where you have a bounding box.

There are many tasks (e.g. image segmentation) where you have an IoU (the predicted segment vs the actual segment), but there are no bounding boxes.


If you are talking about IoBB (Intersection over detected B-Box) used in medical imaging, I think it is different from IoU. IoBB would divide overlapped area over the area of detected B-Box not the Union of it.

In medical imaging, sometimes the detector identified smaller parts of a large or scattered lesion with a big ground-truth bounding-box so IoBB is used.

Actually I am also looking for a strict definition of IoBB since when I calculated the IoU from my result, it was quite different from the reference result where used IoBB instead of IoU..


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.