A Web-based Identification Method for Illegal Streaming Videos Using Low-frequency Components of the Fast Fourier Transform

With the proliferation of web-based content platforms, the distribution of illegally streamed videos poses a serious threat to the reliability of web applications and the integrity of content copyright protection systems. Traditional video identification methods typically require the processing of large-scale feature data, which hinders the real-time performance, lightweight nature, and scalability demanded by web environments. In this paper, we propose a method for identifying illegally streamed videos that is optimized for efficient operation within web systems. The proposed approach utilizes only the low-frequency components of the fast Fourier transform (FFT). By transforming video frames into the frequency domain and extracting the structurally significant low-frequency components, the method replaces high-dimensional feature data with more compact representations. This allows the system to maintain low computational complexity and fast response times, even in web application environments. Experimental results demonstrate that, compared to existing methods, the proposed technique achieves up to 93 times reduction in feature data size, a recognition rate of 98%, and an average response time of 1745 ms. From the perspective of web engineering, the proposed method holds strong potential as a real-time identification module in web-based copyright protection systems. It offers a balanced approach that satisfies both lightweight processing requirements and high accuracy.