This program is an attempt to find the robot's camera pose (and the robot's orientation from that) in a 3D space using two vision targets of known dimension and position on a fixed, veritical 2D plane.
List of requirements:
- DONE: Edge and corner detection of the vision target rectangles on a single test image
- TODO: Get camera image from IP camera (http? more research needed)
- TODO: Use homologic transforms to figure out relative translation of viewed vision targets to real vision targets
- TODO: Use camera calibration matrix and homologic math to map homography H to to camera pose K[R|t] see link for info on this method
- TODO: Open tunnel to cRIO via Google Protocol Buffers
- NICE: Use Google Protocol Buffers to move camera to seek vision targets when outside of range
- TODO: Use Google Protocol Buffers to inform cRIO of camera and/or robot pose relative to vision targets (and the goals, by extension)