Spline Interpolated Edge Detection of Satellite Imagery

asked 2016-05-12 14:14:40 -0500

WxBug36 gravatar image

updated 2016-05-15 13:55:22 -0500

I have never done anything with computer vision before, so forgive me a little bit here.

I am a meteorologist with an interest in computer science, and I would like to experiment with computer vision and visible satellite imagery along the California coast. Here's what I'd like to do.

First, I'd grab the latest satellite imagery (fog band during night, visible during the day) which would look like this: image description

Next, I'd want to run an edge detection (if possible) that would return something like this (green line): image description

My questions are:

  1. Is this even possible or am I way off track here?

  2. Would edge detection (or something similar) work with a complex "edge" that has a large number of points in the polygon such as the image above?

  3. Is spline interpolation possible after edges are detected?

  4. Is it possible to save the detection feature (ie the green line) to an image or geospatial file after the detection is complete for later use?

Thank you!

edit retag flag offensive close merge delete

Comments

Yes it's possible. you must have some knowledge in C++ java or python...

LBerger gravatar imageLBerger ( 2016-05-12 15:43:51 -0500 )edit

I do. I would probably use python. Which cv function/library would help most with this problem?

WxBug36 gravatar imageWxBug36 ( 2016-05-12 15:44:59 -0500 )edit

I don't know python. You should download opencv and try many examples and read docs about example. There is also some books.

LBerger gravatar imageLBerger ( 2016-05-13 01:05:57 -0500 )edit

Were you able to find the answer to your question? I am struggling with a similar problem (wanting to spline an edge)

lash81 gravatar imagelash81 ( 2018-01-03 10:04:27 -0500 )edit