DOI: 10.1109/ICIP.2004.1419400
关键词: Speech recognition 、 Transform coding 、 Watermark 、 White noise 、 Quantization (signal processing) 、 Data compression 、 Digital watermarking 、 Algorithm 、 Lossy compression 、 Mathematics 、 Vector quantization
摘要: In this paper, we introduce a new robust watermarking technique, which is called two-level quantization based spread-spectrum (TQSS). method, the selected transform coefficients of cover signal are quantized to predefined threshold, determined by estimated bound distortion. The quantizer again used for watermark retrieval. Thus, essentially embedded into polarities coefficients. Our simulation results showed that TQSS outperformed some common (SS) methods under variety attacks, such as lossy compression, additive Gaussian white noise, filtering, scaling and print-and-scan process.