音訊信號的隨機性程度測試

Abstract

中文摘要

程式之發展足以讓人類完成許多錯綜複雜以及反覆之演算，許多演算之能力是我們無法理用紙筆進行，必須仰賴程式完成，甚至許多生理訊號的處理更是如此，像是：生理電訊號、生理機械訊號、聲音訊號…等，是完全無法紙上談兵的，因此突顯出程之重要性。 本課程當中之測試隨機訊號，正如同上述所訴之生理訊號，利用程式的輔助將完整訊號讀入，並利用隨機之概念，將測試隨機錄製之訊號是否真正屬於隨機訊號，最終利用課本所學之隨機概念，將程式所繪製之訊號帶入公式，並驗證是否滿足隨機之特性。 利用自己所錄製之語音訊號，將之呈現於程式當中，再利用轉折點(Turning points)的概念，找出與音訊號中所有的峰谷值，並以固定大小之窗框計數每筆時間點窗框內中的峰谷值個數，並了解窗框之大小對語音訊號之影響，以及是否滿足隨機信號(randomness)之特性。

Abstract The development of programs is sufficient for humans to complete many intricate and repetitive calculations. The ability of many calculations is that we can’t do it with pen and paper. We must rely on the program to complete, even the processing of many physiological signals, such as: physiological electrical signals, physiological machinery Signals, sound signals... etc. are completely impossible to talk about on paper, so the importance of the journey is highlighted. The test random signal in this course is just like the physiological signal described above. Use the aid of the program to read in the complete signal, and use the concept of randomness to test whether the randomly recorded signal is truly a random signal, and finally use what you learned in the textbook The concept of randomness is to bring the signal drawn by the program into the formula and verify whether it satisfies the characteristics of randomness. Use the voice signal you recorded to present it in the program, and then use the concept of turning points to find all the peaks and valleys in the audio signal, and count each time point window with a fixed-size window frame The number of peaks and valleys in the frame, and understand the influence of the size of the window frame on the voice signal, and whether it meets the characteristics of randomness.