Revision history [back]

Actually, I have a big doubt about this atritcle. The Internal Energy give the force that perpendicular to the contour, so only the growing term has effect. And this term is suppressed or enhanced by other unrelated forces，the final result will be confusing.

One possible method might be: 1. Grow the ends of the contour by some pixels with the most gradient 2. Run the ACM method to adjust the growed contour to the right place

Actually, I have a big doubt about this atritcle. The Internal Energy give the force that perpendicular to the contour, so only the growing term has effect. And this term is suppressed or enhanced by other unrelated forces，the final result will be confusing.

One possible method might be: 1. Grow the ends of the contour by some pixels with the most gradient gradient. 2. Run the ACM method to adjust the growed contour to the right placeplace.

Actually, I have a big doubt about this atritcle. The Internal Energy give the force that perpendicular to the contour, so only the growing term has effect. And this term is suppressed or enhanced by other unrelated forces，the final result will be confusing.

One possible method might be: 1. Grow the ends of the contour by some pixels with the most gradient. 2. Run the ACM method to adjust the growed contour to the right place.

But still: 1.The growing direction and amount is hard to determine. If only grow by one pixel, after ACM, it might be growed by nothing. 2.It is very easy to grow the contour to the neighbor contour, but they should not be connected actually.

The growing force and growing dircetion of the method in the article is very unclear. if only simply make the growing fore proportional to the gradient magnitude, and add it to the ACM equation，the result is nothing growing.growing. If it does growing ，it is the "luck" do its job.