Abstract:
The battleground Players and their general environment is not quite as same as the
conventional games like (Football, Cricket). Similarly, they face exceptional
difficulties while playing on the Basketball court. So there should be research done to
address these difficulties while carrying out Basketball Referees Hand Sign. The
objective ofthis project is to develop image recognition algorithms to recognize hand
signs in basketball game. This report explores different techniques used for the
recognition of hand signs. Different stages involving image processing like the pre processing stage, segmentation and feature extraction will be studied and discussed.
Finally the end product ofthe algorithms will be written in the software called Jupiter
Notebook. This project uses the Tensor flow object detection API. The main advantage
of using these technique is that they provides features extraction and detection that is
suitable for person recognition. There are multiple of detections in our project for
referee detection we are detecting the referee with whistle and we are successfully
detecting the referee with accuracy ranging between 80 to 90%. This Project report
discusses the most suitable deep-learning models for real time object detection and
recognition and evaluates on the detection and recognition ofthree Referee Signs. The
results of this Project are discussed in previous heading, along with answers to the
research questions formulated in Design and Methodology, some concluding remarks
and future work.
