Orthomosaic vectorization using coddy edge detection [closed]
Dear all, I am MSc student in digital photogrammetry, I have an orthophoto (as shown below) needs to generate a line map (as shown in the free hand sketch) using edge detection filters like Conny, Sobel and Laplacian.
I am civil engineer don't have any knowledge of python, therefore, could anyone help and instruct me how the edge detection can perform?
Good news: It's quite easy to do in OpenCV, and really straightforward. You have functions for edge detection (you'll probably need the Canny detector and HoughLinesP; check the docs).
Bad news: You'll need some Python and OpenCV knowledge. Otherwise you won't be able to get far. But hey, it's not so complicated, and there are lots of books and tutorials!
Dear Kbarni Thank you a lot, could you please send an examples similar to my situation? I am top urgent therefore could you please send some links about such tutorials.
This is a good tutorial on setting up Python with OpenCV. At the end you'll find tutorials to basic image handling (reading, displaying) and simple operations, like edge detection: http://www.bogotobogo.com/python/Open...
python V2.7.5 has been installed with Arc Map, all I need is performing with this version of python OR need additional programs?
Why, in my honest opinion, people with no programming experience try to use software like OpenCV, which is pure coding, I really don't get. Is it manditory that you program something yourself? Or do you just need the result map?
Take a look at these links, which are non-programming solutions
Good luck!
Dear Steven Thank you for comment, actually I don't need to be a programmer, I only need to generate a result map nothing more.
Steven Is it possible to utilizing OpenCV demonstrator (GUI) as a full version? if yes how? because its is shows detected edges only without saving capability.
Well the tool has a edge example, but for OpenCV as it is, you will always need a basic programming and C++ knowledge to get something done. That is why I would urge you to look for any visual tools, which are more intuitive if you have no experience in this.
A new post on the OpenCV homepage might be exactly what you need!