I'm trying to train a model to do image segmentation using Yolov7. I understand the labels are specified in the polygon structure similar to the Coco dataset.
My question is when the polygon data is specified, is there an assumption that the first point of the polygon connects to the last point of the polygon to "close" it? Or is there no such assumption? In which case, a polygon would become effectively a line?
My second question is if I do want to specify a line for image segmentation would it be enough to specify a polygon as (x1,y1) (x2,y2), (x1,y1)? Or do I need to create some fake width to the line for better detection?
