Given 2 markers in space, system can classify and detect each of them in RGB images. Then calculating their 3D positions based on depth maps.
This is project based on Trunk Support Trainer(Trust). In order to develop a 3D detection and localization system that is cheaper than VICON and still maintain high precision. Combining 2-classes YOLO and Kinect v2 depth camera.