基于深度学习的电视图像静帧的识别与监测技术研究  

Research on Recognition and Monitoring Technology of Television Image Static Frames Based on Deep Learning

作　　者：唐崇彦 Tang Chongyan(Guangdong Radio and Television Technology Monitoring Center,Guangdong 510066,China)

机构地区：[1]广东省广播电视技术监测中心,广东510066

出　　处：《广播与电视技术》2024年第6期122-126,共5页Radio & TV Broadcast Engineering

摘　　要：本文提出了一种基于深度学习的电视图像静帧识别与监测方法。该方法以Pytorch深度学习框架为基础,采用swin-Transformer模型对电视图像静帧数据进行不同场景的训练、验证和测试。经过一系列的实验验证,本研究证实了深度学习技术在识别与监测电视图像静帧方面的可行性。在安全播出监管应用中,该方法的引入有望显著提升监测效率,为电视节目的安全播出提供有力保障。This paper proposes a method for recognizing and monitoring static frames in television images based on deep learning.This method is based on the Pytorch deep learning framework and uses the Swin Transformer model to train,validate,and test TV image still frame data in different scenarios.The feasibility of using deep learning technology to identify and monitor static frames in television images is verified,which can improve monitoring efficiency and provide strong guarantees for the safe broadcasting of television programs.

关 键 词：深度学习 Swin-Transformer 图像识别 监测监管 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]