How to make object detection dataset

[vxgu86]

Object detection dataset need a bounding box around the object instead of segmentation mask.
Any hint to get this bounding box straight out?
Thanks

[msb336]

A quick google search for "image mask to bounding box" returned this tutorial

The basic idea is that if you know the pixel area of your classifier max, it is relatively easy to convert into a box by finding the two pixels the represent the bounding corners of the bounding box:
p(min_x, min_y), p(max_x,max_y)

[catproof]

Object detection dataset need a bounding box around the object instead of segmentation mask.
Any hint to get this bounding box straight out?
Thanks

@vxgu86 did you get this working? maybe you can create a brief tutorial?