Pengembangan Model Deep Learning Untuk Deteksi Nomor Punggung Pemain Sepak Bola Pada Video Tracklet

Putra, Muhammad Firdaus Anvi (2025) Pengembangan Model Deep Learning Untuk Deteksi Nomor Punggung Pemain Sepak Bola Pada Video Tracklet. Other thesis, Institut Teknologi Sepuluh Nopember.

Text 5024211067-Undergraduate_Thesis.pdf - Accepted Version Restricted to Repository staff only Download (8MB) | Request a copy

Abstract

Penelitian ini berfokus pada pengembangan model deep learning untuk mendeteksi nomor punggung pemain sepak bola dalam video tracklet. Mengingat semakin meningkatnya kebutuhan akan sistem otomatis dalam analisis olahraga dan siaran, kemampuan untuk mengidentifikasi nomor punggung secara akurat dalam waktu nyata menjadi sangat penting. Penelitian ini memanfaatkan video tracklet yang diambil dari dataset SoccerNet, yang menghadapi tantangan seperti resolusi rendah, motion blur, dan visibilitas nomor punggung yang hanya muncul dalam sebagian kecil frame. Model deep learning yang diusulkan bertujuan untuk mengatasi tantangan ini dengan secara efektif mengekstrak fitur dari data video untuk mengenali nomor punggung yang berkisar antara 1 hingga 99, serta memberikan label khusus untuk kasus di mana nomor punggung tidak terlihat. Kinerja model dievaluasi berdasarkan akurasi, presisi, recall, dan F1-score, dengan tujuan untuk meningkatkan pengalaman menonton bagi penggemar sepak bola dan memperbaiki efisiensi pelacakan pemain selama pertandingan.
======================================================================================================================================
This study focuses on the development of a deep learning model for detecting soccer player jersey numbers in video tracklets. Given the increasing demand for automated systems in sports analytics and broadcasting, the ability to accurately identify jersey numbers in real-time is essential. The research utilizes video tracklets sourced from the SoccerNet dataset, which presents challenges such as low resolution, motion blur, and visibility of jersey numbers appearing in only a limited number of frames. The proposed deep learning model aims to address these challenges by effectively extracting features from video data to recognize jersey numbers ranging from 1 to 99, as well as providing a special label for instances where the jersey number is not visible. The model's performance is evaluated based on accuracy, precision, recall, and F1-score, with the goal of enhancing the viewing experience for soccer fans and improving the efficiency of player tracking during matches.

Item Type:	Thesis (Other)
Uncontrolled Keywords:	Deep Learning, Soccer Player Jersey Number Detection, Video Tracklet, Machine Learning ================================================================================================================================== Deep Learning, Deteksi Nomor Punggung Pemain, Video Tracklet
Subjects:	T Technology > TA Engineering (General). Civil engineering (General) > TA1637 Image processing--Digital techniques. Image analysis--Data processing.
Divisions:	Faculty of Electrical Technology > Computer Engineering > 90243-(S1) Undergraduate Thesis
Depositing User:	Muhammad Firdaus Anvi Putra
Date Deposited:	31 Jul 2025 06:52
Last Modified:	31 Jul 2025 06:52
URI:	http://repository.its.ac.id/id/eprint/124909

Actions (login required)

View Item