Paper

Code

This program achieves CCC (contrasting concentric circle) tracking by thresholding each frame of the input video using Otsu's method to create a binary image.  A binary complement image is created, and all connected component centroid locations are calculated.  By comparing centroid pixel coordinate locations in the original binary image and its complement, CCC targets can be located since they will have centroids in nearly the exact same position in both the binary image and its complement (because the circles are concentric).  

To achieve orientation tracking, each CCC target is assigned an integer identifier.  This number is associated with the target in the first frame of the video.  In later frames, as CCC targets are located, their positions are compared with those from the previous frame.  It is assumed that each target will not move a significant distance between frames, and therefore each target is assigned the identifier that had the closest location in the previous frame.  Because the distance between CCC targets on the paper in the image is known, by solving the Perspective n-Points problem, the program outputs the translation and rotation matrices that correspond to the translation and rotation of the paper with printed CCC targets.  These matrices are displayed.  

Code

To achieve ArUco (Augmented Reality University of Cordoba) tag tracking, each frame of the input video is read in and thresholded using Otsu's method.  Additionally, an adaptive histogram equalization was used to further improve tracking as this particular input video has poor lighting in some frames.  Using the thresholded binary image, the ArUco tag is detected and its orientation is calculated as each tag has a canonical orientation.

The pyramid coordinates are defined in their own coordinate system, which is then converted to pixel coordinates using the following procedure.  First, the pyramid coordinates are transformed into the coordinate system of the ArUco tag using the detected pose of the tag.  Then, these coordinates are transformed into pixel coordinates using given camera properties such as focal length and image size.  Using these pixel coordinates, the lines connecting the corners of the pyramid are drawn.  

Code