I am using opencv to detect shapes, I have to detect a shape like this
this will be a hand drawn picture, and i am using kinect to detect the picture by checking on frames. Now I get 6 points for these kind of pictures, by removing noise, edge detection, finding contours and than approxDP, now I am not able to detect the shape is right faced, left face, down or up. As the points location change for each picture.
So I want to know given two points, how can I check if there is a line segment between the points or not. After getting there two points, I can calculate from remaining points position that where is the shape pointing.
If anyone has better solution, please share.