Abstract:
This project aims at sphericity testing of footballs of circumference ranging from 68-
70 cm set by the FIFA standards. The CSM (Circumference Sphericity Measurement) machine
used for performing sphericity test of FIFA Footballs takes about 9 minutes and 42 seconds to
test a single football sample. Using Machine Vision approach to perform the test would reduce
the testing time and design expenditure. The implementation of Image Processing techniques
such as Hough Transform and Contour Marking methods would be utilized to enable the
system to record radius readings for sphericity calculation using high resolution camera instead
of traditional sensor being used in CSM Machine. This would result in significance reduction
of cost involved in developing the mechanical design and procuring the traditional sensor for
replication of already available machine (CSM Machine). The improvement in measurement
will be achieved using Artificial Intelligence techniques such as Convolutional Neural Network
and Holistically-Nested Edge Detection. The testing time for a single football sample would
be reduced to approximately 1 minute. The machine can be controlled with a Windows
compatible software providing the graphical analysis of the measurements as well
