<![CDATA[mage recognition of sports equipment is often carried out for detection purposes, balls as sports equipment are detected in the form of round objects for various purposes such as goalkeeping robots, ball collection, or monitoring the position of the ball in a match. Classification of various types of sports balls with various shapes located in various areas in one image has not been widely done. This research aims to classify several sports balls, namely soccer balls, hockey balls and shuttlecocks in images with various backgrounds where the images used are images taken randomly from Google Images. Pre-processing in this research starts from changing the image size, changing the GRB image to gray and removing noise using a Gaussian filter. Feature extraction in this research uses two different methods, namely Hough to get line features and LBP (Local Binary Pattern) to get texture features. The feature extraction results were then classified using several classification algorithms with the highest results using Random Forest and LBP feature extraction, namely 70% and the lowest accuracy results using KNN with LBP feature extraction at 43%.]]>
Copyrights © 2024
