Published Paper Details:

In recent years, with the rapid development of Artificial Intelligence technology, the emergence of intelligent digital video surveillance systems, real-time sports competition playback and other technologies have made the advantages of computer vision tasks based on deep learning more and more obvious. It has high flexibility and openness, which indicates the development direction of intelligent image processing. So here, a proposed model is represented for football pose detection using machine learning algorithm to identify and detect the posture of a football player. The system has two phases: first to extract the data points data from the image using the Media Pipe pose estimation library and the second phase is pre-processing the obtained data, training and testing the data using classification-based machine learning algorithms. The proposed system works on 3 football player postures. And a GUI-based desktop application using the Tkinter library is developed. The input is pre-processed in the form of a live video, the object is detected, and the core of the human body is identified using the Media Pipe and OpenCV library. For training and testing data, the stored data which is trained from live video of various postures is used. The logistic regression models for training and testing is used. The system gets an accuracy score of 90%. This developed model is useful in fields like learning players posture and the probability of goal , for training the players by studying the opponents team player's posture , coaches can use to improve their player's performance by observing and training them through their posture