Reputation: 419
Hai I am developing a robot for shortest path finding. This system consists of
the mobile robot, trajectory planner, motion controller, visual
sensor (CCD camera), ZigBee wireless communication device
and a maze terrain. The camera is used to capture images of
the mobile robot within the maze. Developed image processing
and analyzing algorithms determine the robot’s position and
orientation based on color markers recognition.I am developing its software part in java.
My aim is to create a graph of conected nodes from captured image.And apply shortest path algorthom (BFS) to find the shortest path.In this black line indicates path and Red indicates conected nodes.
I have written a program to convert image to a matrix. marked white pixels as 0,black 1,red 2. My plan is to estimate the conecting nodes from this matrix. How i will be able to do that .? any alternative idea ? Please help me
Upvotes: 2
Views: 994
Reputation: 7230
Well, basically, you would have to vectorize the image first; there's another post dicussing how to do this
Once the image is converted to vectors, it should be should be relatively less difficult to convert the vectors to a graph structure.
-- edit --
It just occurred to me that it won't be necessary to convert your matrix to a graph: The A* algorithm could be applied directly to the matrix.
Upvotes: 1