I'm working on a project where I want to get height of object in image.
I'm able to identify the object and get bounding box using object detection but the box is not precise(little bigger). Can I crop the image with bounding box coordinates and then apply some mechanism like edge detection or something to get exact bounding box or height.
Since semantic segmentation is heavier I don't wanna go ahead with that also my trained models are in yolo.
Edit: Added pic with the object detected- object is stop sign as u can see the bounding box is little bigger with respect to height..