Thresholding in Canny edge detection algorithm

Since Canny recommended a ratio of high:low threshold to be between 2:1 and 3:1. I am interested in knowing why is this range particularly selected? Does the answer lie within the understanding of how humans look at images? Any link or explanation would be much appreciated.

Answer 1

[Image Analyst]

I would guess it's just a rule of thumb based on general experience - that's why it's such a wide range. You'll have to fine tune the number for your particular kind of images.