In this paper we develop a non-intrusive method for detecting vital signs in a passenger vehicle. Various types of sensors can detect vital signs, including body temperature, carbon dioxide, and body vibrations. Velocity sensors that are convenient and accurate at acquiring data are adopted in this research. Studies have shown that the human body generates vertical vibrations at a low frequency, mainly caused by an involuntary vibration phenomenon, called "ballistocardiac vibration". This vibration can be detected with sensors at a low frequency range. If people hide in a vehicle, vital vibrations are transmitted to the body of the vehicle, making it resonate at the same frequency range. These vibration signals offer an effective method for detecting people concealed in a vehicle. To increase the system's accuracy, signal processing is essential for extracting a specific vital-frequency band, by calculating the signals of a person in a vehicle from several velocity sensors.