| Abstract [eng] |
The aim of this work is to create a system for recognizing gestures of the basketball referee, using different combinations of feature extraction and classification algorithms, and to analyze the created system. In the first section of the final work, a review of the literature was carried out in order to determine the methods for pre-processing, distinguishing features and classification suitable for the gesture recognition task. In the second part the developed algorithm of the basketball referee sign recognition system is reviewed. In the third chapter, investigations of the created system are carried out. The most important parameters of the implemented system are the number of mistakenly detected commands, accuracy, classification training time and classification time. |
