使用二次離散餘弦變換隱藏資料於音訊之初步研究

Abstract

由於現在智慧型手機及平板電腦的普及和其效能的提升，我們提出將資訊嵌入音訊的方式來傳遞並由這些高效能的手持式裝置進行即時擷取，儘管這樣傳遞資訊的方式有傳輸的流量和距離不高的缺點以及較容易會受到環境噪音的影響，但比起類似功能的方式，如二維條碼，此方式不但具有可多人同時存取的優點且設備需求不高。不同於常見的展頻浮水印方式，我們提出基於離散餘弦轉換將資訊隱藏在音訊頻帶的方法，並透過錯誤控制碼的技術降低擷取資訊的錯誤率，進而達到不錯的傳輸量以及容錯性，且在一般的噪音環境下仍有良好的表現。

Owing to the growing popularity and enhancing of smartphones and tablets, we propose a method embedding information into audio signal to transmit and extracted by portable devices. Though this method is constrained by low data throughput and not long transmission distance and influenced by environmental noise easily, it can be access by multi-user simultaneously comparing to the similar way such as two-dimensional barcode and its device requirements are not high. We hide information into specified frequency bands of audio signal based on DCT differently from the traditional spread-spectrum method. And we using error control code to lower error rate of extracted information. Therefore, our method can achieve feasible transmission throughput and robustness within common noisy environments.