Computer Vision: Non-Max Suppression (Object Detection)

A bounding box is a rectangle around an object that specifies its location in the image, its class and a probability metric of confidence describing how likely the object is within that box.

We either use the corner points (x1, y1, x2, y2) or length dimensions along with the center of the bounding box (xc, yc, Height, Width) to enumerate the location of the bounding box.

It is represented using a list-data structure as follows:

• b_box = [x1, y1, x2, y2, class, confidence]