Custom AR Tag Detection, Tracking & Manipulation
Using OpenCV in Python
Project Description
The following project uses OpenCV to detect custom AR tags in a video, track them, obtain information from them and project 2D and 3D objects on them.
Introduction
An AR Tag is nothing but an easily recognizable point of reference in an Augmented Reality application which helps to calibrate and track camera position in a scene. These markers generally look like QR code sequence and have Tag ID, orientation information and, to make them easy to recognize black padding attached to the outer layer.
The camera when reads these tags in the image can know its position and orientation with respect to the marker and hence know its own location in the scene (world coordinates)
Marker Description
The markers we will be detecting and tracking in this project are custom markers which are made up of 8x8 boxes as shown in the image below. The outer most layer, which consists of 2 blocks of black, is the padding layer. This is added so that tag detection is easy in the frame of the video. The layer of single blocks just inside this padding decides the orientation of the tag. When the white block is located at the bottom right, it…
