IIC Journal of Innovation 10th Edition | Página 17

Intelligent Realities For Workers Using Augmented Reality, Virtual Reality and Beyond
chess play. 22 It should be possible to use
computer vision to interpret and analyze a
chess position on a physical board. This
capability could be packaged in an app that
aids the parent volunteer in both ensuring
legal chess play and providing chess
coaching and knowledge.
headsets are both expensive and new to
most potential parent volunteers. Smart
phones, on the other hand, are already in the
pockets of most parents.
The computer vision problem breaks in to
four parts – finding the board in the image,
creating a 3D coordinate space that finds all
64 squares, recognizing the pieces in legal
play on the board, and then correctly placing
the pieces on the squares. Then the position
can be stated in the standardized Forsyth-
Edwards Notation (FEN) and passed to a
chess analysis engine. The engine can then
check if the position is legal, checkmate,
draw or stalemate and communicate that to
the volunteer. It can also analyze the
position and provide coaching info. The
architecture is illustrated in Figure 4.
But the economics of youth chess clubs are
daunting. Chess equipment is cheap, with a
set that will last twenty years costing cents
per player per year. While there are sensor-
laden boards that can stream moves across
the Internet in the IoT style where the
“things” are the chess pieces, such as the
DGT smart board, they are much too
expensive for a typical club. Computer vision
is the better economic choice over a sensor
approach.
Either a smart phone or an AR head set is a
reasonable choice to host the app. But
22
M. Thomas, “Scholastic Chess Clubs: 10 Reasons Why,” SAS Voices, Aug 2014 [online]. Available:
https://blogs.sas.com/content/sascom/2014/08/29/scholastic-chess-clubs-10-reasons-why/
- 13 -
March 2019