The Intervention of Digital Media Technology on the Visual Language of Radio and Television

Abstract

Abstract This paper first explores the changes brought about by the convergence of digital media technology and radio and television. The SIFT feature extraction method is proposed based on the image signal characteristics, and the convolution operation is done on both the extracted image and the Gaussian kernel. In order to ensure the accurate localization of the key points of the image, the method of quadratic Taylor expansion of the DoG function is utilized to remove the unstable points, thus obtaining the accurate localization of the key points. Next, a 128-dimensional SIFT feature vector is generated by specifying the direction reference of the key point, and a visual dictionary is constructed on the basis of the generated feature vector. The impact of digital media technology on language intervention in radio and television is investigated through quantitative analysis and case studies. The results show that the scores of the image features after SIFT processing are in the interval of [13.342, 15.636], which is the highest score among the three methods, and the effect is better. The global digital media news parameter works increased by 240 works from 2013-2014 and reached the peak state of 732 works in 2021, and the public’s recognition of the visual language visualization of digital media news increased.