Speech signal is a widely-used type of information, with the development of artificial intelligence, the processing algorithms of speech signals is also rapidly evolving. The key to speech signal processing is to extract the feature information of speech signals. This paper extracts the fundamental period and LPC parameters of speech, and detects the resonance peak information of speech signals by focusing on real-time collected speech signals and related preprocessing. Based on understanding the principles of typical algorithms, the programming implementation and algorithmic results are provided, which can be applied to many places where speech signals are used, such as speech compression, speech recognition, and speech synthesis.