Computer Vision

Computer Vision can help robots perform better in the competition by utilizing an aimbot. The aimbot operates by receiving images from a high-speed industrial camera and the images feed through a series of algorithmic processing to obtain the desired features from the target armor plates. The desired features will translate to the information that the embedded need to control the gimbal of the robot for aiming.

We are experimenting with two types of computer vision pipeline. The first one is the classical detection algorithm through OpenCV, which provides a high-speed detection rate but lower robustness. The second pipeline implements machine learning algorithms and different neural network models. We will be testing different models to find the one that works best with the armor detection task and also ways to optimize and increase the speed of the pipeline.