Computer Vision Simulation for FRC SteamWorks Gear Lift Vision Targets. Included are OpenGL Simulation, Machine Learning Algorithm to map 2D coordinates (picture) into 3D coordinates (real world), and a simulation of the path the robot will take based on the outputs of the Machine Learning algorithm.

To understand the purpose and execution of this project please refer to the following blog posts:

1. Blog #1

2. Blog #2

3. Blog #3

At the end of my mentorship I gave a brief talk to the Chicago Python Group. Video

/OpenGL/Simulation.ipynb Contains the simulation that writes data to a CSV database so we can train the Neural Network. Basically puts a series of points that define the target in 3D space. It them simulates a robot moving around and converging upon this target.

/OpenGL/Train Neural Network.ipynb Loads the generated datasets and trains a Neural Network. These are then saved to a file so they can be used later.

/OpenGL/Simulation and Prediction.ipynb Combines the Simulation with the Neural Network and draws a best path to the target. The angle generated is the one in which the robot will travel.

/Practice ML Contains practice for the Machine Learning algorithms I've been working on.

/Robot Contains an implementation of this that may work on our robot.

OpenGL - simulation

OpenCV - to capture images from robot

scikit-learn - train a neural network

pygame - create window to display simulation

numpy, pandas, csv, etc - data processing