Here we are providing the python and arduino code which we used in Pixelate 20, an image processsing event. It is organised under Technex, an annual Technical fest of IIT(BHU).
Complete problem statement is here. Basically we have to design a BOT which can detect different shape and colour and can find an optimal path described in PS. And can carry boxex from one point to another by using that path.
-
Here we are attaching the feed which our bot gets from the camera placed above the arena.
-
We used OpenCV for shape detection and colour detection. A mask image of yellow coloured object is here-
-
Then find there coordinates and make a 9X9 matrix from these shapes. Here is an image where the black and white dots showing the 81 coordinates of coloured object which are stored in a 9X9 matrix-
-
Then an optimal path was plotted to get the optimal trajectory(we used dijkstra theorem here).
-
Once we have path,we know the coodinates of source and destination and also we used Aruco markers to find the coordinates of our bot. Here is an image of 5X5 aruco-
- Here is the video of our working bot.