| As China’s international status continues to improve,the dissemination of Chinese national instrumental music,a treasure of the country’s rich history and culture,has achieved unprecedented growth.In today’s rapidly developing era of the Internet,the issue of how to strengthen copyright protection for Chinese national instrumental music works while widely disseminating them is a pressing concern.Audio digital watermarking technology has emerged as an important solution due to its high concealment and excellent security features.After analyzing and summarizing existing copyright protection algorithms based on audio digital watermarking technology,this thesis has designed a zero-watermark copyright protection algorithm for Chinese ethnic instrumental music works,tailored to their unique characteristics.The work includes the following aspects:(1)This thesis has analyzed and summarized existing audio digital watermarking algorithms and designed a zero-watermark copyright protection scheme suitable for Chinese ethnic instrumental music works based on their unique characteristics,which distinguish them from other music genres.The scheme selects melody,chord,and timbre as the key features of ethnic instrumental music works.It mainly comprises three parts: a preprocessing system for ethnic instrumental music,a zero-watermark construction system,and a zero-watermark detection system.(2)To preprocess national instrumental music works,this thesis has designed an overall processing model and planned the organizational relationship of each functional module.To better reflect the characteristics of national instrumental music,a new key frame filtering algorithm based on the principle of general audio key frame filtering is proposed.The thesis also provides a detailed discussion of the algorithm design and implementation for key functional modules,including framing and windowing,audio noise reduction,and linear scaling,which are essential components of the preprocessing system for ethnic instrumental music works.(3)This thesis has proposed a zero-watermark construction scheme for national instrumental music works,based on their unique melody,chord,and timbre characteristics.The scheme is divided into two stages:feature extraction and watermark generation,with a processing model designed for each stage.The thesis plans the functional modules and organizational relationships implemented by the model.A new melody extraction algorithm for ethnic instrumental music works is proposed using candidate pitch sets and dynamic programming.A new chord feature extraction algorithm for instrumental music works is designed using the constant Q transform and principal component analysis to reduce dimensions.Additionally,the thesis provides a detailed algorithm design and implementation for key functional modules,including voice feature extraction,coordination factor calculation,feature binarization,and fusion processing in zero-watermark construction.(4)This thesis has presented a zero-watermark detection scheme for national instrumental music works,with a processing model designed for the scheme and functional modules and organizational relationships planned for their implementation.The determination of copyright ownership is achieved through similarity calculation.To improve the accuracy of copyright judgment,a multi-factor similarity calculation method is proposed.The thesis also provides a detailed algorithm design and implementation for key modules,including the secondary construction of national instrumental music features,watermark feature information analysis,and copyright judgment in zero-watermark detection.(5)This thesis has presented an experimental analysis and verification of the designed zero-watermark copyright protection scheme for ethnic instrumental music works in terms of its usability,imperceptibility,and robustness.The experimental results demonstrate that the scheme has strong pertinence and good antiattack performance,indicating its effectiveness in protecting ethnic instrumental music works from unauthorized use and distribution. |