RESEARCH ON ACTION DECOMPOSITION METHOD OF FOOTBALL VIDEO IMAGE BASED ON DEEP LEARNING
Haoyu Wang
Sports Science and Technology College, Guangzhou Institute of Applied Science and Technology, 526072, China
Lanfeng Wang
Sports College, Guizhou City Vocational College, Guiyang, 550025, China
Litong Wang
Gdansk University of Physical Education and Sport, Gdansk, 80-336, Poland
Xinran Wang
Shandong Culture and Tourism Investment Group Financial Leasing Co., Ltd, 250000, China
Abstract
In order to better master football skills, a football video image action decomposition method based on deep learning is proposed. Combined with the principle of deep learning, the action characteristics of football video image are collected, the denoising algorithm of football video action image is optimized, and the action decomposition steps of football video image are simplified. Finally, the experiment proves that the football video image action decomposition method based on deep learning has high practicability in the process of practical application, fully meet the research requirements.
Keywords: Deep Learning; Football Video Image; Action Decomposition